diff --git "a/checkpoint-610/trainer_state.json" "b/checkpoint-610/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-610/trainer_state.json" @@ -0,0 +1,4567 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.6003937007874016, + "eval_steps": 153, + "global_step": 610, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.000984251968503937, + "grad_norm": NaN, + "learning_rate": 0.0, + "loss": 6.0688, + "step": 1 + }, + { + "epoch": 0.001968503937007874, + "grad_norm": NaN, + "learning_rate": 0.0, + "loss": 7.5576, + "step": 2 + }, + { + "epoch": 0.002952755905511811, + "grad_norm": Infinity, + "learning_rate": 0.0, + "loss": 4.6849, + "step": 3 + }, + { + "epoch": 0.003937007874015748, + "grad_norm": Infinity, + "learning_rate": 0.0, + "loss": 5.4503, + "step": 4 + }, + { + "epoch": 0.004921259842519685, + "grad_norm": 32.947513580322266, + "learning_rate": 9.940357852882705e-10, + "loss": 5.6057, + "step": 5 + }, + { + "epoch": 0.005905511811023622, + "grad_norm": 40.39633560180664, + "learning_rate": 1.988071570576541e-09, + "loss": 6.3049, + "step": 6 + }, + { + "epoch": 0.006889763779527559, + "grad_norm": 44.38965606689453, + "learning_rate": 2.9821073558648116e-09, + "loss": 6.8336, + "step": 7 + }, + { + "epoch": 0.007874015748031496, + "grad_norm": 28.46092987060547, + "learning_rate": 3.976143141153082e-09, + "loss": 5.0777, + "step": 8 + }, + { + "epoch": 0.008858267716535433, + "grad_norm": 25.817373275756836, + "learning_rate": 4.970178926441353e-09, + "loss": 4.8358, + "step": 9 + }, + { + "epoch": 0.00984251968503937, + "grad_norm": 27.06571388244629, + "learning_rate": 5.964214711729623e-09, + "loss": 4.641, + "step": 10 + }, + { + "epoch": 0.010826771653543307, + "grad_norm": 27.857297897338867, + "learning_rate": 6.9582504970178946e-09, + "loss": 4.828, + "step": 11 + }, + { + "epoch": 0.011811023622047244, + "grad_norm": 29.35377311706543, + "learning_rate": 7.952286282306164e-09, + "loss": 5.2269, + "step": 12 + }, + { + "epoch": 0.012795275590551181, + "grad_norm": 32.06682586669922, + "learning_rate": 8.946322067594435e-09, + "loss": 5.6772, + "step": 13 + }, + { + "epoch": 0.013779527559055118, + "grad_norm": 30.774206161499023, + "learning_rate": 9.940357852882705e-09, + "loss": 5.1422, + "step": 14 + }, + { + "epoch": 0.014763779527559055, + "grad_norm": 45.917659759521484, + "learning_rate": 1.0934393638170978e-08, + "loss": 6.2469, + "step": 15 + }, + { + "epoch": 0.015748031496062992, + "grad_norm": 30.189363479614258, + "learning_rate": 1.1928429423459246e-08, + "loss": 4.6802, + "step": 16 + }, + { + "epoch": 0.01673228346456693, + "grad_norm": 25.023223876953125, + "learning_rate": 1.2922465208747517e-08, + "loss": 4.5492, + "step": 17 + }, + { + "epoch": 0.017716535433070866, + "grad_norm": 26.336748123168945, + "learning_rate": 1.3916500994035789e-08, + "loss": 4.8062, + "step": 18 + }, + { + "epoch": 0.018700787401574805, + "grad_norm": 51.350032806396484, + "learning_rate": 1.4910536779324056e-08, + "loss": 7.5141, + "step": 19 + }, + { + "epoch": 0.01968503937007874, + "grad_norm": 33.953224182128906, + "learning_rate": 1.590457256461233e-08, + "loss": 5.5202, + "step": 20 + }, + { + "epoch": 0.02066929133858268, + "grad_norm": 36.828433990478516, + "learning_rate": 1.68986083499006e-08, + "loss": 6.5025, + "step": 21 + }, + { + "epoch": 0.021653543307086614, + "grad_norm": 43.68119812011719, + "learning_rate": 1.789264413518887e-08, + "loss": 7.318, + "step": 22 + }, + { + "epoch": 0.022637795275590553, + "grad_norm": 23.9952392578125, + "learning_rate": 1.888667992047714e-08, + "loss": 4.6458, + "step": 23 + }, + { + "epoch": 0.023622047244094488, + "grad_norm": 19.315523147583008, + "learning_rate": 1.988071570576541e-08, + "loss": 4.6191, + "step": 24 + }, + { + "epoch": 0.024606299212598427, + "grad_norm": 22.747207641601562, + "learning_rate": 2.087475149105368e-08, + "loss": 4.3159, + "step": 25 + }, + { + "epoch": 0.025590551181102362, + "grad_norm": 32.90699768066406, + "learning_rate": 2.1868787276341955e-08, + "loss": 6.3677, + "step": 26 + }, + { + "epoch": 0.0265748031496063, + "grad_norm": 25.224443435668945, + "learning_rate": 2.2862823061630224e-08, + "loss": 5.6052, + "step": 27 + }, + { + "epoch": 0.027559055118110236, + "grad_norm": 16.32752799987793, + "learning_rate": 2.3856858846918493e-08, + "loss": 4.196, + "step": 28 + }, + { + "epoch": 0.028543307086614175, + "grad_norm": 19.50617218017578, + "learning_rate": 2.4850894632206765e-08, + "loss": 4.4802, + "step": 29 + }, + { + "epoch": 0.02952755905511811, + "grad_norm": 21.613880157470703, + "learning_rate": 2.5844930417495034e-08, + "loss": 4.9193, + "step": 30 + }, + { + "epoch": 0.03051181102362205, + "grad_norm": 16.059247970581055, + "learning_rate": 2.6838966202783303e-08, + "loss": 4.0996, + "step": 31 + }, + { + "epoch": 0.031496062992125984, + "grad_norm": 27.45536994934082, + "learning_rate": 2.7833001988071578e-08, + "loss": 5.6307, + "step": 32 + }, + { + "epoch": 0.03248031496062992, + "grad_norm": 19.249103546142578, + "learning_rate": 2.8827037773359847e-08, + "loss": 4.5745, + "step": 33 + }, + { + "epoch": 0.03346456692913386, + "grad_norm": 17.516521453857422, + "learning_rate": 2.982107355864811e-08, + "loss": 4.4514, + "step": 34 + }, + { + "epoch": 0.0344488188976378, + "grad_norm": 11.684857368469238, + "learning_rate": 3.081510934393639e-08, + "loss": 4.0617, + "step": 35 + }, + { + "epoch": 0.03543307086614173, + "grad_norm": 16.813453674316406, + "learning_rate": 3.180914512922466e-08, + "loss": 5.0298, + "step": 36 + }, + { + "epoch": 0.03641732283464567, + "grad_norm": 10.92259407043457, + "learning_rate": 3.280318091451293e-08, + "loss": 3.9815, + "step": 37 + }, + { + "epoch": 0.03740157480314961, + "grad_norm": 11.449631690979004, + "learning_rate": 3.37972166998012e-08, + "loss": 4.0871, + "step": 38 + }, + { + "epoch": 0.038385826771653545, + "grad_norm": 11.223074913024902, + "learning_rate": 3.479125248508947e-08, + "loss": 4.2378, + "step": 39 + }, + { + "epoch": 0.03937007874015748, + "grad_norm": 9.864290237426758, + "learning_rate": 3.578528827037774e-08, + "loss": 3.8226, + "step": 40 + }, + { + "epoch": 0.040354330708661415, + "grad_norm": 11.084328651428223, + "learning_rate": 3.6779324055666005e-08, + "loss": 4.3519, + "step": 41 + }, + { + "epoch": 0.04133858267716536, + "grad_norm": 9.965168952941895, + "learning_rate": 3.777335984095428e-08, + "loss": 3.6345, + "step": 42 + }, + { + "epoch": 0.04232283464566929, + "grad_norm": 18.375577926635742, + "learning_rate": 3.8767395626242556e-08, + "loss": 5.0829, + "step": 43 + }, + { + "epoch": 0.04330708661417323, + "grad_norm": 11.232723236083984, + "learning_rate": 3.976143141153082e-08, + "loss": 4.6701, + "step": 44 + }, + { + "epoch": 0.04429133858267716, + "grad_norm": 10.896126747131348, + "learning_rate": 4.0755467196819094e-08, + "loss": 4.1371, + "step": 45 + }, + { + "epoch": 0.045275590551181105, + "grad_norm": 10.237566947937012, + "learning_rate": 4.174950298210736e-08, + "loss": 4.2418, + "step": 46 + }, + { + "epoch": 0.04625984251968504, + "grad_norm": 13.484391212463379, + "learning_rate": 4.274353876739563e-08, + "loss": 4.4766, + "step": 47 + }, + { + "epoch": 0.047244094488188976, + "grad_norm": 9.769365310668945, + "learning_rate": 4.373757455268391e-08, + "loss": 4.4797, + "step": 48 + }, + { + "epoch": 0.04822834645669291, + "grad_norm": 9.953036308288574, + "learning_rate": 4.4731610337972176e-08, + "loss": 3.8471, + "step": 49 + }, + { + "epoch": 0.04921259842519685, + "grad_norm": 15.378759384155273, + "learning_rate": 4.572564612326045e-08, + "loss": 4.3194, + "step": 50 + }, + { + "epoch": 0.05019685039370079, + "grad_norm": 9.474905967712402, + "learning_rate": 4.6719681908548713e-08, + "loss": 3.9426, + "step": 51 + }, + { + "epoch": 0.051181102362204724, + "grad_norm": 8.882713317871094, + "learning_rate": 4.7713717693836986e-08, + "loss": 3.5333, + "step": 52 + }, + { + "epoch": 0.05216535433070866, + "grad_norm": 14.956167221069336, + "learning_rate": 4.870775347912525e-08, + "loss": 4.2426, + "step": 53 + }, + { + "epoch": 0.0531496062992126, + "grad_norm": 9.333856582641602, + "learning_rate": 4.970178926441353e-08, + "loss": 3.9816, + "step": 54 + }, + { + "epoch": 0.054133858267716536, + "grad_norm": 9.034503936767578, + "learning_rate": 5.06958250497018e-08, + "loss": 3.663, + "step": 55 + }, + { + "epoch": 0.05511811023622047, + "grad_norm": 10.114209175109863, + "learning_rate": 5.168986083499007e-08, + "loss": 3.9057, + "step": 56 + }, + { + "epoch": 0.05610236220472441, + "grad_norm": 9.802759170532227, + "learning_rate": 5.268389662027834e-08, + "loss": 4.0345, + "step": 57 + }, + { + "epoch": 0.05708661417322835, + "grad_norm": 10.378804206848145, + "learning_rate": 5.3677932405566605e-08, + "loss": 3.5233, + "step": 58 + }, + { + "epoch": 0.058070866141732284, + "grad_norm": 10.237386703491211, + "learning_rate": 5.467196819085488e-08, + "loss": 3.7999, + "step": 59 + }, + { + "epoch": 0.05905511811023622, + "grad_norm": 11.103493690490723, + "learning_rate": 5.5666003976143156e-08, + "loss": 3.1885, + "step": 60 + }, + { + "epoch": 0.060039370078740155, + "grad_norm": 11.967391014099121, + "learning_rate": 5.666003976143142e-08, + "loss": 3.6013, + "step": 61 + }, + { + "epoch": 0.0610236220472441, + "grad_norm": 10.79468822479248, + "learning_rate": 5.7654075546719694e-08, + "loss": 3.392, + "step": 62 + }, + { + "epoch": 0.06200787401574803, + "grad_norm": 13.18323802947998, + "learning_rate": 5.864811133200796e-08, + "loss": 3.3814, + "step": 63 + }, + { + "epoch": 0.06299212598425197, + "grad_norm": 15.541842460632324, + "learning_rate": 5.964214711729623e-08, + "loss": 4.0428, + "step": 64 + }, + { + "epoch": 0.0639763779527559, + "grad_norm": 13.166780471801758, + "learning_rate": 6.06361829025845e-08, + "loss": 3.7825, + "step": 65 + }, + { + "epoch": 0.06496062992125984, + "grad_norm": 12.48865795135498, + "learning_rate": 6.163021868787278e-08, + "loss": 3.4181, + "step": 66 + }, + { + "epoch": 0.06594488188976377, + "grad_norm": 14.15877628326416, + "learning_rate": 6.262425447316104e-08, + "loss": 3.7793, + "step": 67 + }, + { + "epoch": 0.06692913385826772, + "grad_norm": 18.16546630859375, + "learning_rate": 6.361829025844931e-08, + "loss": 3.8344, + "step": 68 + }, + { + "epoch": 0.06791338582677166, + "grad_norm": 14.758286476135254, + "learning_rate": 6.461232604373759e-08, + "loss": 3.2165, + "step": 69 + }, + { + "epoch": 0.0688976377952756, + "grad_norm": 15.924538612365723, + "learning_rate": 6.560636182902586e-08, + "loss": 3.3811, + "step": 70 + }, + { + "epoch": 0.06988188976377953, + "grad_norm": 17.181415557861328, + "learning_rate": 6.660039761431412e-08, + "loss": 3.5984, + "step": 71 + }, + { + "epoch": 0.07086614173228346, + "grad_norm": 16.691272735595703, + "learning_rate": 6.75944333996024e-08, + "loss": 3.8583, + "step": 72 + }, + { + "epoch": 0.0718503937007874, + "grad_norm": 17.478809356689453, + "learning_rate": 6.858846918489067e-08, + "loss": 3.296, + "step": 73 + }, + { + "epoch": 0.07283464566929133, + "grad_norm": 15.149248123168945, + "learning_rate": 6.958250497017893e-08, + "loss": 2.7661, + "step": 74 + }, + { + "epoch": 0.07381889763779527, + "grad_norm": 15.139701843261719, + "learning_rate": 7.057654075546721e-08, + "loss": 2.9805, + "step": 75 + }, + { + "epoch": 0.07480314960629922, + "grad_norm": 14.794472694396973, + "learning_rate": 7.157057654075548e-08, + "loss": 2.566, + "step": 76 + }, + { + "epoch": 0.07578740157480315, + "grad_norm": 15.433120727539062, + "learning_rate": 7.256461232604374e-08, + "loss": 3.258, + "step": 77 + }, + { + "epoch": 0.07677165354330709, + "grad_norm": 22.558115005493164, + "learning_rate": 7.355864811133201e-08, + "loss": 3.3804, + "step": 78 + }, + { + "epoch": 0.07775590551181102, + "grad_norm": 19.630340576171875, + "learning_rate": 7.455268389662029e-08, + "loss": 2.8828, + "step": 79 + }, + { + "epoch": 0.07874015748031496, + "grad_norm": 17.1998233795166, + "learning_rate": 7.554671968190855e-08, + "loss": 3.1077, + "step": 80 + }, + { + "epoch": 0.0797244094488189, + "grad_norm": 13.535666465759277, + "learning_rate": 7.654075546719683e-08, + "loss": 2.9441, + "step": 81 + }, + { + "epoch": 0.08070866141732283, + "grad_norm": 14.031355857849121, + "learning_rate": 7.753479125248511e-08, + "loss": 2.9465, + "step": 82 + }, + { + "epoch": 0.08169291338582677, + "grad_norm": 13.701343536376953, + "learning_rate": 7.852882703777338e-08, + "loss": 2.7088, + "step": 83 + }, + { + "epoch": 0.08267716535433071, + "grad_norm": 15.102946281433105, + "learning_rate": 7.952286282306164e-08, + "loss": 2.9215, + "step": 84 + }, + { + "epoch": 0.08366141732283465, + "grad_norm": 20.15928077697754, + "learning_rate": 8.051689860834992e-08, + "loss": 3.4698, + "step": 85 + }, + { + "epoch": 0.08464566929133858, + "grad_norm": 14.807561874389648, + "learning_rate": 8.151093439363819e-08, + "loss": 2.2414, + "step": 86 + }, + { + "epoch": 0.08562992125984252, + "grad_norm": 18.390995025634766, + "learning_rate": 8.250497017892645e-08, + "loss": 3.1601, + "step": 87 + }, + { + "epoch": 0.08661417322834646, + "grad_norm": 17.47426986694336, + "learning_rate": 8.349900596421472e-08, + "loss": 2.7714, + "step": 88 + }, + { + "epoch": 0.08759842519685039, + "grad_norm": 19.336660385131836, + "learning_rate": 8.4493041749503e-08, + "loss": 3.0311, + "step": 89 + }, + { + "epoch": 0.08858267716535433, + "grad_norm": 22.308195114135742, + "learning_rate": 8.548707753479126e-08, + "loss": 3.0336, + "step": 90 + }, + { + "epoch": 0.08956692913385826, + "grad_norm": 16.85112762451172, + "learning_rate": 8.648111332007953e-08, + "loss": 1.9358, + "step": 91 + }, + { + "epoch": 0.09055118110236221, + "grad_norm": 18.357120513916016, + "learning_rate": 8.747514910536782e-08, + "loss": 2.6031, + "step": 92 + }, + { + "epoch": 0.09153543307086615, + "grad_norm": 21.082101821899414, + "learning_rate": 8.846918489065609e-08, + "loss": 2.7515, + "step": 93 + }, + { + "epoch": 0.09251968503937008, + "grad_norm": 21.116485595703125, + "learning_rate": 8.946322067594435e-08, + "loss": 2.8496, + "step": 94 + }, + { + "epoch": 0.09350393700787402, + "grad_norm": 15.967782020568848, + "learning_rate": 9.045725646123262e-08, + "loss": 1.8015, + "step": 95 + }, + { + "epoch": 0.09448818897637795, + "grad_norm": 20.430376052856445, + "learning_rate": 9.14512922465209e-08, + "loss": 2.8138, + "step": 96 + }, + { + "epoch": 0.09547244094488189, + "grad_norm": 18.769359588623047, + "learning_rate": 9.244532803180916e-08, + "loss": 2.0597, + "step": 97 + }, + { + "epoch": 0.09645669291338582, + "grad_norm": 19.360021591186523, + "learning_rate": 9.343936381709743e-08, + "loss": 2.1053, + "step": 98 + }, + { + "epoch": 0.09744094488188976, + "grad_norm": 22.48013687133789, + "learning_rate": 9.44333996023857e-08, + "loss": 2.6785, + "step": 99 + }, + { + "epoch": 0.0984251968503937, + "grad_norm": 21.50940704345703, + "learning_rate": 9.542743538767397e-08, + "loss": 2.588, + "step": 100 + }, + { + "epoch": 0.09940944881889764, + "grad_norm": 17.409393310546875, + "learning_rate": 9.642147117296224e-08, + "loss": 2.0099, + "step": 101 + }, + { + "epoch": 0.10039370078740158, + "grad_norm": 22.641996383666992, + "learning_rate": 9.74155069582505e-08, + "loss": 2.7947, + "step": 102 + }, + { + "epoch": 0.10137795275590551, + "grad_norm": 18.91701889038086, + "learning_rate": 9.840954274353878e-08, + "loss": 2.3274, + "step": 103 + }, + { + "epoch": 0.10236220472440945, + "grad_norm": 21.317251205444336, + "learning_rate": 9.940357852882706e-08, + "loss": 2.2545, + "step": 104 + }, + { + "epoch": 0.10334645669291338, + "grad_norm": 24.298248291015625, + "learning_rate": 1.0039761431411533e-07, + "loss": 2.4575, + "step": 105 + }, + { + "epoch": 0.10433070866141732, + "grad_norm": 19.284252166748047, + "learning_rate": 1.013916500994036e-07, + "loss": 2.4413, + "step": 106 + }, + { + "epoch": 0.10531496062992125, + "grad_norm": 19.619426727294922, + "learning_rate": 1.0238568588469187e-07, + "loss": 2.3185, + "step": 107 + }, + { + "epoch": 0.1062992125984252, + "grad_norm": 20.323917388916016, + "learning_rate": 1.0337972166998014e-07, + "loss": 2.1577, + "step": 108 + }, + { + "epoch": 0.10728346456692914, + "grad_norm": 20.885499954223633, + "learning_rate": 1.043737574552684e-07, + "loss": 2.1278, + "step": 109 + }, + { + "epoch": 0.10826771653543307, + "grad_norm": 21.687036514282227, + "learning_rate": 1.0536779324055668e-07, + "loss": 2.0967, + "step": 110 + }, + { + "epoch": 0.10925196850393701, + "grad_norm": 23.031898498535156, + "learning_rate": 1.0636182902584495e-07, + "loss": 2.6142, + "step": 111 + }, + { + "epoch": 0.11023622047244094, + "grad_norm": 20.64027976989746, + "learning_rate": 1.0735586481113321e-07, + "loss": 1.8553, + "step": 112 + }, + { + "epoch": 0.11122047244094488, + "grad_norm": 18.268877029418945, + "learning_rate": 1.0834990059642149e-07, + "loss": 2.1523, + "step": 113 + }, + { + "epoch": 0.11220472440944881, + "grad_norm": 22.190549850463867, + "learning_rate": 1.0934393638170976e-07, + "loss": 2.1726, + "step": 114 + }, + { + "epoch": 0.11318897637795275, + "grad_norm": 18.314851760864258, + "learning_rate": 1.1033797216699802e-07, + "loss": 1.8564, + "step": 115 + }, + { + "epoch": 0.1141732283464567, + "grad_norm": 19.17725944519043, + "learning_rate": 1.1133200795228631e-07, + "loss": 1.8413, + "step": 116 + }, + { + "epoch": 0.11515748031496063, + "grad_norm": 24.518449783325195, + "learning_rate": 1.1232604373757458e-07, + "loss": 2.0441, + "step": 117 + }, + { + "epoch": 0.11614173228346457, + "grad_norm": 24.78040313720703, + "learning_rate": 1.1332007952286284e-07, + "loss": 2.2159, + "step": 118 + }, + { + "epoch": 0.1171259842519685, + "grad_norm": 28.552289962768555, + "learning_rate": 1.1431411530815111e-07, + "loss": 2.6779, + "step": 119 + }, + { + "epoch": 0.11811023622047244, + "grad_norm": 21.049219131469727, + "learning_rate": 1.1530815109343939e-07, + "loss": 2.2976, + "step": 120 + }, + { + "epoch": 0.11909448818897637, + "grad_norm": 23.962575912475586, + "learning_rate": 1.1630218687872765e-07, + "loss": 1.9407, + "step": 121 + }, + { + "epoch": 0.12007874015748031, + "grad_norm": 20.15193748474121, + "learning_rate": 1.1729622266401592e-07, + "loss": 1.9019, + "step": 122 + }, + { + "epoch": 0.12106299212598425, + "grad_norm": 22.563459396362305, + "learning_rate": 1.182902584493042e-07, + "loss": 2.2149, + "step": 123 + }, + { + "epoch": 0.1220472440944882, + "grad_norm": 16.86638832092285, + "learning_rate": 1.1928429423459245e-07, + "loss": 1.6823, + "step": 124 + }, + { + "epoch": 0.12303149606299213, + "grad_norm": 20.249584197998047, + "learning_rate": 1.2027833001988073e-07, + "loss": 1.8402, + "step": 125 + }, + { + "epoch": 0.12401574803149606, + "grad_norm": 19.5487117767334, + "learning_rate": 1.21272365805169e-07, + "loss": 1.6914, + "step": 126 + }, + { + "epoch": 0.125, + "grad_norm": 19.093505859375, + "learning_rate": 1.222664015904573e-07, + "loss": 2.1626, + "step": 127 + }, + { + "epoch": 0.12598425196850394, + "grad_norm": 17.598743438720703, + "learning_rate": 1.2326043737574557e-07, + "loss": 1.6414, + "step": 128 + }, + { + "epoch": 0.12696850393700787, + "grad_norm": 22.712181091308594, + "learning_rate": 1.2425447316103382e-07, + "loss": 2.2043, + "step": 129 + }, + { + "epoch": 0.1279527559055118, + "grad_norm": 20.516571044921875, + "learning_rate": 1.2524850894632207e-07, + "loss": 1.9987, + "step": 130 + }, + { + "epoch": 0.12893700787401574, + "grad_norm": 22.58023452758789, + "learning_rate": 1.2624254473161035e-07, + "loss": 1.8868, + "step": 131 + }, + { + "epoch": 0.12992125984251968, + "grad_norm": 17.800390243530273, + "learning_rate": 1.2723658051689863e-07, + "loss": 1.8262, + "step": 132 + }, + { + "epoch": 0.1309055118110236, + "grad_norm": 20.7684383392334, + "learning_rate": 1.282306163021869e-07, + "loss": 2.0404, + "step": 133 + }, + { + "epoch": 0.13188976377952755, + "grad_norm": 20.88973617553711, + "learning_rate": 1.2922465208747519e-07, + "loss": 1.9134, + "step": 134 + }, + { + "epoch": 0.1328740157480315, + "grad_norm": 21.78728485107422, + "learning_rate": 1.3021868787276344e-07, + "loss": 2.3725, + "step": 135 + }, + { + "epoch": 0.13385826771653545, + "grad_norm": 16.885374069213867, + "learning_rate": 1.3121272365805172e-07, + "loss": 1.4127, + "step": 136 + }, + { + "epoch": 0.13484251968503938, + "grad_norm": 18.944265365600586, + "learning_rate": 1.3220675944333997e-07, + "loss": 1.6876, + "step": 137 + }, + { + "epoch": 0.13582677165354332, + "grad_norm": 18.579692840576172, + "learning_rate": 1.3320079522862825e-07, + "loss": 1.8376, + "step": 138 + }, + { + "epoch": 0.13681102362204725, + "grad_norm": 19.68845558166504, + "learning_rate": 1.3419483101391653e-07, + "loss": 1.6992, + "step": 139 + }, + { + "epoch": 0.1377952755905512, + "grad_norm": 21.27842903137207, + "learning_rate": 1.351888667992048e-07, + "loss": 1.5032, + "step": 140 + }, + { + "epoch": 0.13877952755905512, + "grad_norm": 28.506824493408203, + "learning_rate": 1.3618290258449306e-07, + "loss": 2.0334, + "step": 141 + }, + { + "epoch": 0.13976377952755906, + "grad_norm": 24.97527313232422, + "learning_rate": 1.3717693836978134e-07, + "loss": 2.3581, + "step": 142 + }, + { + "epoch": 0.140748031496063, + "grad_norm": 22.391145706176758, + "learning_rate": 1.381709741550696e-07, + "loss": 1.4236, + "step": 143 + }, + { + "epoch": 0.14173228346456693, + "grad_norm": 22.977148056030273, + "learning_rate": 1.3916500994035787e-07, + "loss": 2.202, + "step": 144 + }, + { + "epoch": 0.14271653543307086, + "grad_norm": 25.096189498901367, + "learning_rate": 1.4015904572564615e-07, + "loss": 1.7654, + "step": 145 + }, + { + "epoch": 0.1437007874015748, + "grad_norm": 28.528640747070312, + "learning_rate": 1.4115308151093443e-07, + "loss": 1.5748, + "step": 146 + }, + { + "epoch": 0.14468503937007873, + "grad_norm": 30.80350685119629, + "learning_rate": 1.421471172962227e-07, + "loss": 1.7996, + "step": 147 + }, + { + "epoch": 0.14566929133858267, + "grad_norm": 22.726850509643555, + "learning_rate": 1.4314115308151096e-07, + "loss": 1.7517, + "step": 148 + }, + { + "epoch": 0.1466535433070866, + "grad_norm": 24.926170349121094, + "learning_rate": 1.4413518886679924e-07, + "loss": 1.8933, + "step": 149 + }, + { + "epoch": 0.14763779527559054, + "grad_norm": 20.570375442504883, + "learning_rate": 1.451292246520875e-07, + "loss": 1.2836, + "step": 150 + }, + { + "epoch": 0.1486220472440945, + "grad_norm": 21.08650779724121, + "learning_rate": 1.4612326043737577e-07, + "loss": 1.7145, + "step": 151 + }, + { + "epoch": 0.14960629921259844, + "grad_norm": 24.37209701538086, + "learning_rate": 1.4711729622266402e-07, + "loss": 1.6499, + "step": 152 + }, + { + "epoch": 0.15059055118110237, + "grad_norm": 26.801511764526367, + "learning_rate": 1.4811133200795232e-07, + "loss": 1.8273, + "step": 153 + }, + { + "epoch": 0.15059055118110237, + "eval_Qnli-dev_cosine_accuracy": 0.646484375, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8057259321212769, + "eval_Qnli-dev_cosine_ap": 0.6720663622193426, + "eval_Qnli-dev_cosine_f1": 0.6688102893890675, + "eval_Qnli-dev_cosine_f1_threshold": 0.7187118530273438, + "eval_Qnli-dev_cosine_precision": 0.538860103626943, + "eval_Qnli-dev_cosine_recall": 0.8813559322033898, + "eval_Qnli-dev_dot_accuracy": 0.646484375, + "eval_Qnli-dev_dot_accuracy_threshold": 618.8643798828125, + "eval_Qnli-dev_dot_ap": 0.672083506527328, + "eval_Qnli-dev_dot_f1": 0.6688102893890675, + "eval_Qnli-dev_dot_f1_threshold": 552.0260009765625, + "eval_Qnli-dev_dot_precision": 0.538860103626943, + "eval_Qnli-dev_dot_recall": 0.8813559322033898, + "eval_Qnli-dev_euclidean_accuracy": 0.646484375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 17.27533721923828, + "eval_Qnli-dev_euclidean_ap": 0.6720591998758361, + "eval_Qnli-dev_euclidean_f1": 0.6688102893890675, + "eval_Qnli-dev_euclidean_f1_threshold": 20.787063598632812, + "eval_Qnli-dev_euclidean_precision": 0.538860103626943, + "eval_Qnli-dev_euclidean_recall": 0.8813559322033898, + "eval_Qnli-dev_manhattan_accuracy": 0.6484375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 386.58905029296875, + "eval_Qnli-dev_manhattan_ap": 0.6724653688821339, + "eval_Qnli-dev_manhattan_f1": 0.6645569620253164, + "eval_Qnli-dev_manhattan_f1_threshold": 462.609130859375, + "eval_Qnli-dev_manhattan_precision": 0.5303030303030303, + "eval_Qnli-dev_manhattan_recall": 0.8898305084745762, + "eval_Qnli-dev_max_accuracy": 0.6484375, + "eval_Qnli-dev_max_accuracy_threshold": 618.8643798828125, + "eval_Qnli-dev_max_ap": 0.6724653688821339, + "eval_Qnli-dev_max_f1": 0.6688102893890675, + "eval_Qnli-dev_max_f1_threshold": 552.0260009765625, + "eval_Qnli-dev_max_precision": 0.538860103626943, + "eval_Qnli-dev_max_recall": 0.8898305084745762, + "eval_allNLI-dev_cosine_accuracy": 0.67578125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9427558183670044, + "eval_allNLI-dev_cosine_ap": 0.4368751759846574, + "eval_allNLI-dev_cosine_f1": 0.5225225225225225, + "eval_allNLI-dev_cosine_f1_threshold": 0.8046966791152954, + "eval_allNLI-dev_cosine_precision": 0.3795811518324607, + "eval_allNLI-dev_cosine_recall": 0.838150289017341, + "eval_allNLI-dev_dot_accuracy": 0.67578125, + "eval_allNLI-dev_dot_accuracy_threshold": 724.1080322265625, + "eval_allNLI-dev_dot_ap": 0.436842886797982, + "eval_allNLI-dev_dot_f1": 0.5225225225225225, + "eval_allNLI-dev_dot_f1_threshold": 618.074951171875, + "eval_allNLI-dev_dot_precision": 0.3795811518324607, + "eval_allNLI-dev_dot_recall": 0.838150289017341, + "eval_allNLI-dev_euclidean_accuracy": 0.67578125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.377331733703613, + "eval_allNLI-dev_euclidean_ap": 0.4368602200677977, + "eval_allNLI-dev_euclidean_f1": 0.5225225225225225, + "eval_allNLI-dev_euclidean_f1_threshold": 17.321048736572266, + "eval_allNLI-dev_euclidean_precision": 0.3795811518324607, + "eval_allNLI-dev_euclidean_recall": 0.838150289017341, + "eval_allNLI-dev_manhattan_accuracy": 0.677734375, + "eval_allNLI-dev_manhattan_accuracy_threshold": 223.6764373779297, + "eval_allNLI-dev_manhattan_ap": 0.43892484929307635, + "eval_allNLI-dev_manhattan_f1": 0.5239852398523985, + "eval_allNLI-dev_manhattan_f1_threshold": 372.31396484375, + "eval_allNLI-dev_manhattan_precision": 0.38482384823848237, + "eval_allNLI-dev_manhattan_recall": 0.8208092485549133, + "eval_allNLI-dev_max_accuracy": 0.677734375, + "eval_allNLI-dev_max_accuracy_threshold": 724.1080322265625, + "eval_allNLI-dev_max_ap": 0.43892484929307635, + "eval_allNLI-dev_max_f1": 0.5239852398523985, + "eval_allNLI-dev_max_f1_threshold": 618.074951171875, + "eval_allNLI-dev_max_precision": 0.38482384823848237, + "eval_allNLI-dev_max_recall": 0.838150289017341, + "eval_runtime": 1.2344, + "eval_samples_per_second": 0.0, + "eval_sequential_score": 0.6724653688821339, + "eval_steps_per_second": 0.0, + "eval_sts-test_pearson_cosine": 0.3774946012125992, + "eval_sts-test_pearson_dot": 0.3774648453085433, + "eval_sts-test_pearson_euclidean": 0.38652243004790016, + "eval_sts-test_pearson_manhattan": 0.3861982631744407, + "eval_sts-test_pearson_max": 0.38652243004790016, + "eval_sts-test_spearman_cosine": 0.4056589966976888, + "eval_sts-test_spearman_dot": 0.40563469676275316, + "eval_sts-test_spearman_euclidean": 0.4056589966976888, + "eval_sts-test_spearman_manhattan": 0.4059364545183154, + "eval_sts-test_spearman_max": 0.4059364545183154, + "step": 153 + }, + { + "epoch": 0.1515748031496063, + "grad_norm": 27.144521713256836, + "learning_rate": 1.4910536779324058e-07, + "loss": 2.2859, + "step": 154 + }, + { + "epoch": 0.15255905511811024, + "grad_norm": 16.310321807861328, + "learning_rate": 1.5009940357852886e-07, + "loss": 1.0833, + "step": 155 + }, + { + "epoch": 0.15354330708661418, + "grad_norm": 18.788604736328125, + "learning_rate": 1.510934393638171e-07, + "loss": 1.6829, + "step": 156 + }, + { + "epoch": 0.1545275590551181, + "grad_norm": 23.34528923034668, + "learning_rate": 1.5208747514910539e-07, + "loss": 2.1464, + "step": 157 + }, + { + "epoch": 0.15551181102362205, + "grad_norm": 19.06661033630371, + "learning_rate": 1.5308151093439367e-07, + "loss": 1.745, + "step": 158 + }, + { + "epoch": 0.15649606299212598, + "grad_norm": 23.650297164916992, + "learning_rate": 1.5407554671968192e-07, + "loss": 1.7319, + "step": 159 + }, + { + "epoch": 0.15748031496062992, + "grad_norm": 22.01628303527832, + "learning_rate": 1.5506958250497022e-07, + "loss": 1.6968, + "step": 160 + }, + { + "epoch": 0.15846456692913385, + "grad_norm": 20.905855178833008, + "learning_rate": 1.5606361829025848e-07, + "loss": 1.7401, + "step": 161 + }, + { + "epoch": 0.1594488188976378, + "grad_norm": 22.40980339050293, + "learning_rate": 1.5705765407554675e-07, + "loss": 1.729, + "step": 162 + }, + { + "epoch": 0.16043307086614172, + "grad_norm": 23.574447631835938, + "learning_rate": 1.58051689860835e-07, + "loss": 2.0782, + "step": 163 + }, + { + "epoch": 0.16141732283464566, + "grad_norm": 28.43741798400879, + "learning_rate": 1.5904572564612329e-07, + "loss": 2.6545, + "step": 164 + }, + { + "epoch": 0.1624015748031496, + "grad_norm": 18.536827087402344, + "learning_rate": 1.6003976143141154e-07, + "loss": 1.4045, + "step": 165 + }, + { + "epoch": 0.16338582677165353, + "grad_norm": 17.013965606689453, + "learning_rate": 1.6103379721669984e-07, + "loss": 1.2937, + "step": 166 + }, + { + "epoch": 0.1643700787401575, + "grad_norm": 16.021455764770508, + "learning_rate": 1.620278330019881e-07, + "loss": 1.1171, + "step": 167 + }, + { + "epoch": 0.16535433070866143, + "grad_norm": 20.248655319213867, + "learning_rate": 1.6302186878727637e-07, + "loss": 1.3537, + "step": 168 + }, + { + "epoch": 0.16633858267716536, + "grad_norm": 21.731477737426758, + "learning_rate": 1.6401590457256465e-07, + "loss": 1.7028, + "step": 169 + }, + { + "epoch": 0.1673228346456693, + "grad_norm": 20.372398376464844, + "learning_rate": 1.650099403578529e-07, + "loss": 1.4143, + "step": 170 + }, + { + "epoch": 0.16830708661417323, + "grad_norm": 21.89658546447754, + "learning_rate": 1.6600397614314118e-07, + "loss": 1.8648, + "step": 171 + }, + { + "epoch": 0.16929133858267717, + "grad_norm": 24.402984619140625, + "learning_rate": 1.6699801192842944e-07, + "loss": 1.6768, + "step": 172 + }, + { + "epoch": 0.1702755905511811, + "grad_norm": 28.650039672851562, + "learning_rate": 1.6799204771371774e-07, + "loss": 1.9528, + "step": 173 + }, + { + "epoch": 0.17125984251968504, + "grad_norm": 18.586729049682617, + "learning_rate": 1.68986083499006e-07, + "loss": 1.1718, + "step": 174 + }, + { + "epoch": 0.17224409448818898, + "grad_norm": 22.66750717163086, + "learning_rate": 1.6998011928429427e-07, + "loss": 1.8176, + "step": 175 + }, + { + "epoch": 0.1732283464566929, + "grad_norm": 15.157270431518555, + "learning_rate": 1.7097415506958253e-07, + "loss": 0.8439, + "step": 176 + }, + { + "epoch": 0.17421259842519685, + "grad_norm": 23.08325958251953, + "learning_rate": 1.719681908548708e-07, + "loss": 1.5092, + "step": 177 + }, + { + "epoch": 0.17519685039370078, + "grad_norm": 19.215696334838867, + "learning_rate": 1.7296222664015906e-07, + "loss": 1.1947, + "step": 178 + }, + { + "epoch": 0.17618110236220472, + "grad_norm": 23.268930435180664, + "learning_rate": 1.7395626242544734e-07, + "loss": 1.6395, + "step": 179 + }, + { + "epoch": 0.17716535433070865, + "grad_norm": 24.91169548034668, + "learning_rate": 1.7495029821073564e-07, + "loss": 1.4394, + "step": 180 + }, + { + "epoch": 0.1781496062992126, + "grad_norm": 26.92424201965332, + "learning_rate": 1.759443339960239e-07, + "loss": 1.7548, + "step": 181 + }, + { + "epoch": 0.17913385826771652, + "grad_norm": 22.831315994262695, + "learning_rate": 1.7693836978131217e-07, + "loss": 1.1181, + "step": 182 + }, + { + "epoch": 0.18011811023622049, + "grad_norm": 18.39922332763672, + "learning_rate": 1.7793240556660042e-07, + "loss": 1.0271, + "step": 183 + }, + { + "epoch": 0.18110236220472442, + "grad_norm": 30.0064697265625, + "learning_rate": 1.789264413518887e-07, + "loss": 2.3108, + "step": 184 + }, + { + "epoch": 0.18208661417322836, + "grad_norm": 25.531293869018555, + "learning_rate": 1.7992047713717695e-07, + "loss": 2.1242, + "step": 185 + }, + { + "epoch": 0.1830708661417323, + "grad_norm": 25.3813533782959, + "learning_rate": 1.8091451292246523e-07, + "loss": 1.9822, + "step": 186 + }, + { + "epoch": 0.18405511811023623, + "grad_norm": 22.24681282043457, + "learning_rate": 1.819085487077535e-07, + "loss": 2.3605, + "step": 187 + }, + { + "epoch": 0.18503937007874016, + "grad_norm": 19.291316986083984, + "learning_rate": 1.829025844930418e-07, + "loss": 1.5251, + "step": 188 + }, + { + "epoch": 0.1860236220472441, + "grad_norm": 18.280820846557617, + "learning_rate": 1.8389662027833004e-07, + "loss": 1.2351, + "step": 189 + }, + { + "epoch": 0.18700787401574803, + "grad_norm": 18.341115951538086, + "learning_rate": 1.8489065606361832e-07, + "loss": 1.5859, + "step": 190 + }, + { + "epoch": 0.18799212598425197, + "grad_norm": 21.981475830078125, + "learning_rate": 1.8588469184890657e-07, + "loss": 1.8056, + "step": 191 + }, + { + "epoch": 0.1889763779527559, + "grad_norm": 23.482725143432617, + "learning_rate": 1.8687872763419485e-07, + "loss": 1.349, + "step": 192 + }, + { + "epoch": 0.18996062992125984, + "grad_norm": 15.451715469360352, + "learning_rate": 1.8787276341948313e-07, + "loss": 0.893, + "step": 193 + }, + { + "epoch": 0.19094488188976377, + "grad_norm": 25.24738121032715, + "learning_rate": 1.888667992047714e-07, + "loss": 1.5122, + "step": 194 + }, + { + "epoch": 0.1919291338582677, + "grad_norm": 22.348981857299805, + "learning_rate": 1.898608349900597e-07, + "loss": 1.3875, + "step": 195 + }, + { + "epoch": 0.19291338582677164, + "grad_norm": 22.770132064819336, + "learning_rate": 1.9085487077534794e-07, + "loss": 1.29, + "step": 196 + }, + { + "epoch": 0.19389763779527558, + "grad_norm": 31.219194412231445, + "learning_rate": 1.9184890656063622e-07, + "loss": 2.2931, + "step": 197 + }, + { + "epoch": 0.19488188976377951, + "grad_norm": 24.657859802246094, + "learning_rate": 1.9284294234592447e-07, + "loss": 1.2663, + "step": 198 + }, + { + "epoch": 0.19586614173228348, + "grad_norm": 26.825185775756836, + "learning_rate": 1.9383697813121275e-07, + "loss": 1.9712, + "step": 199 + }, + { + "epoch": 0.1968503937007874, + "grad_norm": 30.97873306274414, + "learning_rate": 1.94831013916501e-07, + "loss": 2.3307, + "step": 200 + }, + { + "epoch": 0.19783464566929135, + "grad_norm": 24.922365188598633, + "learning_rate": 1.958250497017893e-07, + "loss": 1.6544, + "step": 201 + }, + { + "epoch": 0.19881889763779528, + "grad_norm": 22.842063903808594, + "learning_rate": 1.9681908548707756e-07, + "loss": 1.638, + "step": 202 + }, + { + "epoch": 0.19980314960629922, + "grad_norm": 17.376962661743164, + "learning_rate": 1.9781312127236584e-07, + "loss": 1.3412, + "step": 203 + }, + { + "epoch": 0.20078740157480315, + "grad_norm": 17.828535079956055, + "learning_rate": 1.9880715705765412e-07, + "loss": 1.4454, + "step": 204 + }, + { + "epoch": 0.2017716535433071, + "grad_norm": 18.73072624206543, + "learning_rate": 1.9980119284294237e-07, + "loss": 1.5437, + "step": 205 + }, + { + "epoch": 0.20275590551181102, + "grad_norm": 22.05230140686035, + "learning_rate": 2.0079522862823065e-07, + "loss": 1.4921, + "step": 206 + }, + { + "epoch": 0.20374015748031496, + "grad_norm": 18.090892791748047, + "learning_rate": 2.017892644135189e-07, + "loss": 1.4298, + "step": 207 + }, + { + "epoch": 0.2047244094488189, + "grad_norm": 22.94427490234375, + "learning_rate": 2.027833001988072e-07, + "loss": 1.6174, + "step": 208 + }, + { + "epoch": 0.20570866141732283, + "grad_norm": 20.903331756591797, + "learning_rate": 2.0377733598409546e-07, + "loss": 1.4137, + "step": 209 + }, + { + "epoch": 0.20669291338582677, + "grad_norm": 22.67809295654297, + "learning_rate": 2.0477137176938374e-07, + "loss": 1.5652, + "step": 210 + }, + { + "epoch": 0.2076771653543307, + "grad_norm": 16.824064254760742, + "learning_rate": 2.05765407554672e-07, + "loss": 1.1631, + "step": 211 + }, + { + "epoch": 0.20866141732283464, + "grad_norm": 21.46413230895996, + "learning_rate": 2.0675944333996027e-07, + "loss": 1.2351, + "step": 212 + }, + { + "epoch": 0.20964566929133857, + "grad_norm": 22.461610794067383, + "learning_rate": 2.0775347912524852e-07, + "loss": 1.7537, + "step": 213 + }, + { + "epoch": 0.2106299212598425, + "grad_norm": 22.150938034057617, + "learning_rate": 2.087475149105368e-07, + "loss": 1.3186, + "step": 214 + }, + { + "epoch": 0.21161417322834647, + "grad_norm": 22.82622718811035, + "learning_rate": 2.097415506958251e-07, + "loss": 1.2258, + "step": 215 + }, + { + "epoch": 0.2125984251968504, + "grad_norm": 13.514169692993164, + "learning_rate": 2.1073558648111336e-07, + "loss": 0.7695, + "step": 216 + }, + { + "epoch": 0.21358267716535434, + "grad_norm": 22.254596710205078, + "learning_rate": 2.1172962226640164e-07, + "loss": 1.2775, + "step": 217 + }, + { + "epoch": 0.21456692913385828, + "grad_norm": 23.31028938293457, + "learning_rate": 2.127236580516899e-07, + "loss": 1.6795, + "step": 218 + }, + { + "epoch": 0.2155511811023622, + "grad_norm": 24.057025909423828, + "learning_rate": 2.1371769383697817e-07, + "loss": 1.2862, + "step": 219 + }, + { + "epoch": 0.21653543307086615, + "grad_norm": 17.363821029663086, + "learning_rate": 2.1471172962226642e-07, + "loss": 1.1723, + "step": 220 + }, + { + "epoch": 0.21751968503937008, + "grad_norm": 22.69175148010254, + "learning_rate": 2.1570576540755473e-07, + "loss": 1.3322, + "step": 221 + }, + { + "epoch": 0.21850393700787402, + "grad_norm": 24.083465576171875, + "learning_rate": 2.1669980119284298e-07, + "loss": 1.7564, + "step": 222 + }, + { + "epoch": 0.21948818897637795, + "grad_norm": 17.775938034057617, + "learning_rate": 2.1769383697813126e-07, + "loss": 1.1071, + "step": 223 + }, + { + "epoch": 0.2204724409448819, + "grad_norm": 18.539440155029297, + "learning_rate": 2.186878727634195e-07, + "loss": 1.2011, + "step": 224 + }, + { + "epoch": 0.22145669291338582, + "grad_norm": 18.234455108642578, + "learning_rate": 2.196819085487078e-07, + "loss": 1.2303, + "step": 225 + }, + { + "epoch": 0.22244094488188976, + "grad_norm": 20.291522979736328, + "learning_rate": 2.2067594433399604e-07, + "loss": 1.212, + "step": 226 + }, + { + "epoch": 0.2234251968503937, + "grad_norm": 15.60973072052002, + "learning_rate": 2.2166998011928432e-07, + "loss": 1.0117, + "step": 227 + }, + { + "epoch": 0.22440944881889763, + "grad_norm": 20.482995986938477, + "learning_rate": 2.2266401590457263e-07, + "loss": 1.1907, + "step": 228 + }, + { + "epoch": 0.22539370078740156, + "grad_norm": 24.126340866088867, + "learning_rate": 2.2365805168986088e-07, + "loss": 2.1293, + "step": 229 + }, + { + "epoch": 0.2263779527559055, + "grad_norm": 21.279150009155273, + "learning_rate": 2.2465208747514916e-07, + "loss": 1.3063, + "step": 230 + }, + { + "epoch": 0.22736220472440946, + "grad_norm": 20.62582778930664, + "learning_rate": 2.256461232604374e-07, + "loss": 1.2841, + "step": 231 + }, + { + "epoch": 0.2283464566929134, + "grad_norm": 20.63606834411621, + "learning_rate": 2.266401590457257e-07, + "loss": 1.3778, + "step": 232 + }, + { + "epoch": 0.22933070866141733, + "grad_norm": 18.099950790405273, + "learning_rate": 2.2763419483101394e-07, + "loss": 1.2242, + "step": 233 + }, + { + "epoch": 0.23031496062992127, + "grad_norm": 18.96858024597168, + "learning_rate": 2.2862823061630222e-07, + "loss": 0.9227, + "step": 234 + }, + { + "epoch": 0.2312992125984252, + "grad_norm": 22.201169967651367, + "learning_rate": 2.296222664015905e-07, + "loss": 1.2221, + "step": 235 + }, + { + "epoch": 0.23228346456692914, + "grad_norm": 26.657480239868164, + "learning_rate": 2.3061630218687878e-07, + "loss": 2.1041, + "step": 236 + }, + { + "epoch": 0.23326771653543307, + "grad_norm": 25.558757781982422, + "learning_rate": 2.3161033797216703e-07, + "loss": 1.3341, + "step": 237 + }, + { + "epoch": 0.234251968503937, + "grad_norm": 22.0543212890625, + "learning_rate": 2.326043737574553e-07, + "loss": 1.0876, + "step": 238 + }, + { + "epoch": 0.23523622047244094, + "grad_norm": 22.258333206176758, + "learning_rate": 2.335984095427436e-07, + "loss": 1.3328, + "step": 239 + }, + { + "epoch": 0.23622047244094488, + "grad_norm": 21.519908905029297, + "learning_rate": 2.3459244532803184e-07, + "loss": 1.2958, + "step": 240 + }, + { + "epoch": 0.2372047244094488, + "grad_norm": 23.345388412475586, + "learning_rate": 2.3558648111332012e-07, + "loss": 1.1522, + "step": 241 + }, + { + "epoch": 0.23818897637795275, + "grad_norm": 25.306140899658203, + "learning_rate": 2.365805168986084e-07, + "loss": 1.7942, + "step": 242 + }, + { + "epoch": 0.23917322834645668, + "grad_norm": 21.840364456176758, + "learning_rate": 2.3757455268389668e-07, + "loss": 1.1325, + "step": 243 + }, + { + "epoch": 0.24015748031496062, + "grad_norm": 26.5495662689209, + "learning_rate": 2.385685884691849e-07, + "loss": 1.6466, + "step": 244 + }, + { + "epoch": 0.24114173228346455, + "grad_norm": 21.928937911987305, + "learning_rate": 2.395626242544732e-07, + "loss": 1.4608, + "step": 245 + }, + { + "epoch": 0.2421259842519685, + "grad_norm": 16.23383903503418, + "learning_rate": 2.4055666003976146e-07, + "loss": 0.6375, + "step": 246 + }, + { + "epoch": 0.24311023622047245, + "grad_norm": 26.922739028930664, + "learning_rate": 2.4155069582504976e-07, + "loss": 2.0177, + "step": 247 + }, + { + "epoch": 0.2440944881889764, + "grad_norm": 23.482742309570312, + "learning_rate": 2.42544731610338e-07, + "loss": 1.2069, + "step": 248 + }, + { + "epoch": 0.24507874015748032, + "grad_norm": 14.740228652954102, + "learning_rate": 2.4353876739562627e-07, + "loss": 0.7639, + "step": 249 + }, + { + "epoch": 0.24606299212598426, + "grad_norm": 20.238948822021484, + "learning_rate": 2.445328031809146e-07, + "loss": 1.3465, + "step": 250 + }, + { + "epoch": 0.2470472440944882, + "grad_norm": 16.91303253173828, + "learning_rate": 2.4552683896620283e-07, + "loss": 1.064, + "step": 251 + }, + { + "epoch": 0.24803149606299213, + "grad_norm": 22.081308364868164, + "learning_rate": 2.4652087475149113e-07, + "loss": 1.3757, + "step": 252 + }, + { + "epoch": 0.24901574803149606, + "grad_norm": 21.756267547607422, + "learning_rate": 2.475149105367794e-07, + "loss": 1.612, + "step": 253 + }, + { + "epoch": 0.25, + "grad_norm": 16.87566566467285, + "learning_rate": 2.4850894632206764e-07, + "loss": 0.7917, + "step": 254 + }, + { + "epoch": 0.25098425196850394, + "grad_norm": 22.13374900817871, + "learning_rate": 2.495029821073559e-07, + "loss": 1.5515, + "step": 255 + }, + { + "epoch": 0.25196850393700787, + "grad_norm": 18.96138572692871, + "learning_rate": 2.5049701789264414e-07, + "loss": 0.799, + "step": 256 + }, + { + "epoch": 0.2529527559055118, + "grad_norm": 17.10991859436035, + "learning_rate": 2.5149105367793245e-07, + "loss": 0.9882, + "step": 257 + }, + { + "epoch": 0.25393700787401574, + "grad_norm": 15.347769737243652, + "learning_rate": 2.524850894632207e-07, + "loss": 1.1814, + "step": 258 + }, + { + "epoch": 0.2549212598425197, + "grad_norm": 11.857462882995605, + "learning_rate": 2.53479125248509e-07, + "loss": 0.6394, + "step": 259 + }, + { + "epoch": 0.2559055118110236, + "grad_norm": 20.00450325012207, + "learning_rate": 2.5447316103379726e-07, + "loss": 1.4756, + "step": 260 + }, + { + "epoch": 0.25688976377952755, + "grad_norm": 12.062335968017578, + "learning_rate": 2.554671968190855e-07, + "loss": 0.5338, + "step": 261 + }, + { + "epoch": 0.2578740157480315, + "grad_norm": 16.912168502807617, + "learning_rate": 2.564612326043738e-07, + "loss": 0.9779, + "step": 262 + }, + { + "epoch": 0.2588582677165354, + "grad_norm": 22.713825225830078, + "learning_rate": 2.5745526838966207e-07, + "loss": 1.5307, + "step": 263 + }, + { + "epoch": 0.25984251968503935, + "grad_norm": 22.88834571838379, + "learning_rate": 2.5844930417495037e-07, + "loss": 1.1213, + "step": 264 + }, + { + "epoch": 0.2608267716535433, + "grad_norm": 18.133310317993164, + "learning_rate": 2.5944333996023857e-07, + "loss": 0.9482, + "step": 265 + }, + { + "epoch": 0.2618110236220472, + "grad_norm": 19.092018127441406, + "learning_rate": 2.604373757455269e-07, + "loss": 0.9599, + "step": 266 + }, + { + "epoch": 0.26279527559055116, + "grad_norm": 25.283647537231445, + "learning_rate": 2.614314115308152e-07, + "loss": 1.4455, + "step": 267 + }, + { + "epoch": 0.2637795275590551, + "grad_norm": 28.519742965698242, + "learning_rate": 2.6242544731610343e-07, + "loss": 1.6496, + "step": 268 + }, + { + "epoch": 0.26476377952755903, + "grad_norm": 15.56734561920166, + "learning_rate": 2.634194831013917e-07, + "loss": 0.7402, + "step": 269 + }, + { + "epoch": 0.265748031496063, + "grad_norm": 16.450769424438477, + "learning_rate": 2.6441351888667994e-07, + "loss": 0.7835, + "step": 270 + }, + { + "epoch": 0.26673228346456695, + "grad_norm": 18.447402954101562, + "learning_rate": 2.6540755467196824e-07, + "loss": 0.7821, + "step": 271 + }, + { + "epoch": 0.2677165354330709, + "grad_norm": 22.634441375732422, + "learning_rate": 2.664015904572565e-07, + "loss": 1.5422, + "step": 272 + }, + { + "epoch": 0.2687007874015748, + "grad_norm": 20.091428756713867, + "learning_rate": 2.6739562624254475e-07, + "loss": 1.0995, + "step": 273 + }, + { + "epoch": 0.26968503937007876, + "grad_norm": 22.89579200744629, + "learning_rate": 2.6838966202783305e-07, + "loss": 1.378, + "step": 274 + }, + { + "epoch": 0.2706692913385827, + "grad_norm": 25.36764144897461, + "learning_rate": 2.693836978131213e-07, + "loss": 1.3562, + "step": 275 + }, + { + "epoch": 0.27165354330708663, + "grad_norm": 15.717765808105469, + "learning_rate": 2.703777335984096e-07, + "loss": 0.7376, + "step": 276 + }, + { + "epoch": 0.27263779527559057, + "grad_norm": 21.739999771118164, + "learning_rate": 2.7137176938369786e-07, + "loss": 1.1678, + "step": 277 + }, + { + "epoch": 0.2736220472440945, + "grad_norm": 21.12672233581543, + "learning_rate": 2.723658051689861e-07, + "loss": 1.2989, + "step": 278 + }, + { + "epoch": 0.27460629921259844, + "grad_norm": 22.614158630371094, + "learning_rate": 2.7335984095427437e-07, + "loss": 1.9559, + "step": 279 + }, + { + "epoch": 0.2755905511811024, + "grad_norm": 18.742107391357422, + "learning_rate": 2.743538767395627e-07, + "loss": 1.1237, + "step": 280 + }, + { + "epoch": 0.2765748031496063, + "grad_norm": 17.969554901123047, + "learning_rate": 2.75347912524851e-07, + "loss": 0.952, + "step": 281 + }, + { + "epoch": 0.27755905511811024, + "grad_norm": 24.01414680480957, + "learning_rate": 2.763419483101392e-07, + "loss": 1.6629, + "step": 282 + }, + { + "epoch": 0.2785433070866142, + "grad_norm": 21.362796783447266, + "learning_rate": 2.773359840954275e-07, + "loss": 1.871, + "step": 283 + }, + { + "epoch": 0.2795275590551181, + "grad_norm": 19.05644989013672, + "learning_rate": 2.7833001988071574e-07, + "loss": 1.5946, + "step": 284 + }, + { + "epoch": 0.28051181102362205, + "grad_norm": 18.40645408630371, + "learning_rate": 2.7932405566600404e-07, + "loss": 1.4456, + "step": 285 + }, + { + "epoch": 0.281496062992126, + "grad_norm": 18.53467559814453, + "learning_rate": 2.803180914512923e-07, + "loss": 1.4085, + "step": 286 + }, + { + "epoch": 0.2824803149606299, + "grad_norm": 18.1446533203125, + "learning_rate": 2.8131212723658055e-07, + "loss": 1.1394, + "step": 287 + }, + { + "epoch": 0.28346456692913385, + "grad_norm": 15.109747886657715, + "learning_rate": 2.8230616302186885e-07, + "loss": 1.0315, + "step": 288 + }, + { + "epoch": 0.2844488188976378, + "grad_norm": 22.63469696044922, + "learning_rate": 2.833001988071571e-07, + "loss": 1.488, + "step": 289 + }, + { + "epoch": 0.2854330708661417, + "grad_norm": 18.841909408569336, + "learning_rate": 2.842942345924454e-07, + "loss": 1.4006, + "step": 290 + }, + { + "epoch": 0.28641732283464566, + "grad_norm": 20.073745727539062, + "learning_rate": 2.852882703777336e-07, + "loss": 0.9237, + "step": 291 + }, + { + "epoch": 0.2874015748031496, + "grad_norm": 18.586259841918945, + "learning_rate": 2.862823061630219e-07, + "loss": 1.163, + "step": 292 + }, + { + "epoch": 0.28838582677165353, + "grad_norm": 23.599292755126953, + "learning_rate": 2.8727634194831017e-07, + "loss": 1.7037, + "step": 293 + }, + { + "epoch": 0.28937007874015747, + "grad_norm": 16.384347915649414, + "learning_rate": 2.8827037773359847e-07, + "loss": 0.8715, + "step": 294 + }, + { + "epoch": 0.2903543307086614, + "grad_norm": 20.26093101501465, + "learning_rate": 2.892644135188867e-07, + "loss": 1.2101, + "step": 295 + }, + { + "epoch": 0.29133858267716534, + "grad_norm": 19.598831176757812, + "learning_rate": 2.90258449304175e-07, + "loss": 1.1179, + "step": 296 + }, + { + "epoch": 0.29232283464566927, + "grad_norm": 22.65572738647461, + "learning_rate": 2.912524850894633e-07, + "loss": 1.3986, + "step": 297 + }, + { + "epoch": 0.2933070866141732, + "grad_norm": 22.658931732177734, + "learning_rate": 2.9224652087475153e-07, + "loss": 1.7068, + "step": 298 + }, + { + "epoch": 0.29429133858267714, + "grad_norm": 16.94671058654785, + "learning_rate": 2.9324055666003984e-07, + "loss": 0.8695, + "step": 299 + }, + { + "epoch": 0.2952755905511811, + "grad_norm": 20.13766860961914, + "learning_rate": 2.9423459244532804e-07, + "loss": 1.3778, + "step": 300 + }, + { + "epoch": 0.296259842519685, + "grad_norm": 16.939062118530273, + "learning_rate": 2.9522862823061634e-07, + "loss": 1.2834, + "step": 301 + }, + { + "epoch": 0.297244094488189, + "grad_norm": 14.732866287231445, + "learning_rate": 2.9622266401590465e-07, + "loss": 0.8123, + "step": 302 + }, + { + "epoch": 0.29822834645669294, + "grad_norm": 23.4341983795166, + "learning_rate": 2.972166998011929e-07, + "loss": 1.6521, + "step": 303 + }, + { + "epoch": 0.2992125984251969, + "grad_norm": 15.455023765563965, + "learning_rate": 2.9821073558648115e-07, + "loss": 1.1064, + "step": 304 + }, + { + "epoch": 0.3001968503937008, + "grad_norm": 15.683412551879883, + "learning_rate": 2.992047713717694e-07, + "loss": 0.9578, + "step": 305 + }, + { + "epoch": 0.30118110236220474, + "grad_norm": 16.594762802124023, + "learning_rate": 3.001988071570577e-07, + "loss": 0.9254, + "step": 306 + }, + { + "epoch": 0.30118110236220474, + "eval_Qnli-dev_cosine_accuracy": 0.6796875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8338999152183533, + "eval_Qnli-dev_cosine_ap": 0.7040139627388139, + "eval_Qnli-dev_cosine_f1": 0.6716697936210132, + "eval_Qnli-dev_cosine_f1_threshold": 0.7830559015274048, + "eval_Qnli-dev_cosine_precision": 0.6026936026936027, + "eval_Qnli-dev_cosine_recall": 0.7584745762711864, + "eval_Qnli-dev_dot_accuracy": 0.6796875, + "eval_Qnli-dev_dot_accuracy_threshold": 640.5870361328125, + "eval_Qnli-dev_dot_ap": 0.7040479892892765, + "eval_Qnli-dev_dot_f1": 0.6716697936210132, + "eval_Qnli-dev_dot_f1_threshold": 601.5180053710938, + "eval_Qnli-dev_dot_precision": 0.6026936026936027, + "eval_Qnli-dev_dot_recall": 0.7584745762711864, + "eval_Qnli-dev_euclidean_accuracy": 0.6796875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.974668502807617, + "eval_Qnli-dev_euclidean_ap": 0.7040139627388139, + "eval_Qnli-dev_euclidean_f1": 0.6716697936210132, + "eval_Qnli-dev_euclidean_f1_threshold": 18.256603240966797, + "eval_Qnli-dev_euclidean_precision": 0.6026936026936027, + "eval_Qnli-dev_euclidean_recall": 0.7584745762711864, + "eval_Qnli-dev_manhattan_accuracy": 0.677734375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 354.99566650390625, + "eval_Qnli-dev_manhattan_ap": 0.7020916363435745, + "eval_Qnli-dev_manhattan_f1": 0.6715328467153285, + "eval_Qnli-dev_manhattan_f1_threshold": 406.69915771484375, + "eval_Qnli-dev_manhattan_precision": 0.5897435897435898, + "eval_Qnli-dev_manhattan_recall": 0.7796610169491526, + "eval_Qnli-dev_max_accuracy": 0.6796875, + "eval_Qnli-dev_max_accuracy_threshold": 640.5870361328125, + "eval_Qnli-dev_max_ap": 0.7040479892892765, + "eval_Qnli-dev_max_f1": 0.6716697936210132, + "eval_Qnli-dev_max_f1_threshold": 601.5180053710938, + "eval_Qnli-dev_max_precision": 0.6026936026936027, + "eval_Qnli-dev_max_recall": 0.7796610169491526, + "eval_allNLI-dev_cosine_accuracy": 0.685546875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.931303083896637, + "eval_allNLI-dev_cosine_ap": 0.4777228569875479, + "eval_allNLI-dev_cosine_f1": 0.5344467640918581, + "eval_allNLI-dev_cosine_f1_threshold": 0.8382494449615479, + "eval_allNLI-dev_cosine_precision": 0.41830065359477125, + "eval_allNLI-dev_cosine_recall": 0.7398843930635838, + "eval_allNLI-dev_dot_accuracy": 0.685546875, + "eval_allNLI-dev_dot_accuracy_threshold": 715.4099731445312, + "eval_allNLI-dev_dot_ap": 0.47773833073270383, + "eval_allNLI-dev_dot_f1": 0.5344467640918581, + "eval_allNLI-dev_dot_f1_threshold": 643.9212646484375, + "eval_allNLI-dev_dot_precision": 0.41830065359477125, + "eval_allNLI-dev_dot_recall": 0.7398843930635838, + "eval_allNLI-dev_euclidean_accuracy": 0.685546875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.273423194885254, + "eval_allNLI-dev_euclidean_ap": 0.477722856987548, + "eval_allNLI-dev_euclidean_f1": 0.5344467640918581, + "eval_allNLI-dev_euclidean_f1_threshold": 15.764041900634766, + "eval_allNLI-dev_euclidean_precision": 0.41830065359477125, + "eval_allNLI-dev_euclidean_recall": 0.7398843930635838, + "eval_allNLI-dev_manhattan_accuracy": 0.685546875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 236.41517639160156, + "eval_allNLI-dev_manhattan_ap": 0.47890307996210313, + "eval_allNLI-dev_manhattan_f1": 0.5341880341880343, + "eval_allNLI-dev_manhattan_f1_threshold": 339.87042236328125, + "eval_allNLI-dev_manhattan_precision": 0.423728813559322, + "eval_allNLI-dev_manhattan_recall": 0.7225433526011561, + "eval_allNLI-dev_max_accuracy": 0.685546875, + "eval_allNLI-dev_max_accuracy_threshold": 715.4099731445312, + "eval_allNLI-dev_max_ap": 0.47890307996210313, + "eval_allNLI-dev_max_f1": 0.5344467640918581, + "eval_allNLI-dev_max_f1_threshold": 643.9212646484375, + "eval_allNLI-dev_max_precision": 0.423728813559322, + "eval_allNLI-dev_max_recall": 0.7398843930635838, + "eval_runtime": 1.2301, + "eval_samples_per_second": 0.0, + "eval_sequential_score": 0.7040479892892765, + "eval_steps_per_second": 0.0, + "eval_sts-test_pearson_cosine": 0.4665812233506641, + "eval_sts-test_pearson_dot": 0.4665227418822839, + "eval_sts-test_pearson_euclidean": 0.47993570552371345, + "eval_sts-test_pearson_manhattan": 0.4795106797562281, + "eval_sts-test_pearson_max": 0.47993570552371345, + "eval_sts-test_spearman_cosine": 0.48884197010310126, + "eval_sts-test_spearman_dot": 0.4886054596866477, + "eval_sts-test_spearman_euclidean": 0.4888477430158208, + "eval_sts-test_spearman_manhattan": 0.487254956120378, + "eval_sts-test_spearman_max": 0.4888477430158208, + "step": 306 + }, + { + "epoch": 0.3021653543307087, + "grad_norm": 17.415273666381836, + "learning_rate": 3.01192842942346e-07, + "loss": 0.7541, + "step": 307 + }, + { + "epoch": 0.3031496062992126, + "grad_norm": 15.18093204498291, + "learning_rate": 3.021868787276342e-07, + "loss": 0.7324, + "step": 308 + }, + { + "epoch": 0.30413385826771655, + "grad_norm": 14.747346878051758, + "learning_rate": 3.031809145129225e-07, + "loss": 0.5974, + "step": 309 + }, + { + "epoch": 0.3051181102362205, + "grad_norm": 23.3188533782959, + "learning_rate": 3.0417495029821077e-07, + "loss": 1.1481, + "step": 310 + }, + { + "epoch": 0.3061023622047244, + "grad_norm": 28.704936981201172, + "learning_rate": 3.051689860834991e-07, + "loss": 1.6179, + "step": 311 + }, + { + "epoch": 0.30708661417322836, + "grad_norm": 26.603492736816406, + "learning_rate": 3.0616302186878733e-07, + "loss": 1.4641, + "step": 312 + }, + { + "epoch": 0.3080708661417323, + "grad_norm": 30.99028968811035, + "learning_rate": 3.071570576540756e-07, + "loss": 1.7185, + "step": 313 + }, + { + "epoch": 0.3090551181102362, + "grad_norm": 21.485858917236328, + "learning_rate": 3.0815109343936384e-07, + "loss": 0.9328, + "step": 314 + }, + { + "epoch": 0.31003937007874016, + "grad_norm": 20.982534408569336, + "learning_rate": 3.0914512922465214e-07, + "loss": 0.742, + "step": 315 + }, + { + "epoch": 0.3110236220472441, + "grad_norm": 24.897323608398438, + "learning_rate": 3.1013916500994045e-07, + "loss": 1.4173, + "step": 316 + }, + { + "epoch": 0.31200787401574803, + "grad_norm": 18.03276252746582, + "learning_rate": 3.1113320079522865e-07, + "loss": 0.7267, + "step": 317 + }, + { + "epoch": 0.31299212598425197, + "grad_norm": 18.415945053100586, + "learning_rate": 3.1212723658051695e-07, + "loss": 0.9494, + "step": 318 + }, + { + "epoch": 0.3139763779527559, + "grad_norm": 23.052213668823242, + "learning_rate": 3.131212723658052e-07, + "loss": 1.5111, + "step": 319 + }, + { + "epoch": 0.31496062992125984, + "grad_norm": 21.024145126342773, + "learning_rate": 3.141153081510935e-07, + "loss": 1.6949, + "step": 320 + }, + { + "epoch": 0.3159448818897638, + "grad_norm": 24.442323684692383, + "learning_rate": 3.151093439363818e-07, + "loss": 1.7562, + "step": 321 + }, + { + "epoch": 0.3169291338582677, + "grad_norm": 16.658889770507812, + "learning_rate": 3.1610337972167e-07, + "loss": 1.2532, + "step": 322 + }, + { + "epoch": 0.31791338582677164, + "grad_norm": 12.664291381835938, + "learning_rate": 3.170974155069583e-07, + "loss": 1.1086, + "step": 323 + }, + { + "epoch": 0.3188976377952756, + "grad_norm": 16.310880661010742, + "learning_rate": 3.1809145129224657e-07, + "loss": 0.7377, + "step": 324 + }, + { + "epoch": 0.3198818897637795, + "grad_norm": 18.340065002441406, + "learning_rate": 3.190854870775349e-07, + "loss": 1.085, + "step": 325 + }, + { + "epoch": 0.32086614173228345, + "grad_norm": 14.548378944396973, + "learning_rate": 3.200795228628231e-07, + "loss": 0.7767, + "step": 326 + }, + { + "epoch": 0.3218503937007874, + "grad_norm": 19.989688873291016, + "learning_rate": 3.210735586481114e-07, + "loss": 1.4441, + "step": 327 + }, + { + "epoch": 0.3228346456692913, + "grad_norm": 16.6571102142334, + "learning_rate": 3.220675944333997e-07, + "loss": 0.8146, + "step": 328 + }, + { + "epoch": 0.32381889763779526, + "grad_norm": 16.355384826660156, + "learning_rate": 3.2306163021868794e-07, + "loss": 0.7403, + "step": 329 + }, + { + "epoch": 0.3248031496062992, + "grad_norm": 19.589183807373047, + "learning_rate": 3.240556660039762e-07, + "loss": 0.8476, + "step": 330 + }, + { + "epoch": 0.3257874015748031, + "grad_norm": 16.7521915435791, + "learning_rate": 3.2504970178926444e-07, + "loss": 0.7323, + "step": 331 + }, + { + "epoch": 0.32677165354330706, + "grad_norm": 22.762434005737305, + "learning_rate": 3.2604373757455275e-07, + "loss": 1.2241, + "step": 332 + }, + { + "epoch": 0.327755905511811, + "grad_norm": 21.42218017578125, + "learning_rate": 3.27037773359841e-07, + "loss": 1.5065, + "step": 333 + }, + { + "epoch": 0.328740157480315, + "grad_norm": 12.703171730041504, + "learning_rate": 3.280318091451293e-07, + "loss": 0.5259, + "step": 334 + }, + { + "epoch": 0.3297244094488189, + "grad_norm": 22.40658187866211, + "learning_rate": 3.2902584493041756e-07, + "loss": 1.3103, + "step": 335 + }, + { + "epoch": 0.33070866141732286, + "grad_norm": 19.746475219726562, + "learning_rate": 3.300198807157058e-07, + "loss": 0.8655, + "step": 336 + }, + { + "epoch": 0.3316929133858268, + "grad_norm": 19.323755264282227, + "learning_rate": 3.310139165009941e-07, + "loss": 0.7575, + "step": 337 + }, + { + "epoch": 0.33267716535433073, + "grad_norm": 24.301485061645508, + "learning_rate": 3.3200795228628237e-07, + "loss": 1.968, + "step": 338 + }, + { + "epoch": 0.33366141732283466, + "grad_norm": 20.01712417602539, + "learning_rate": 3.330019880715706e-07, + "loss": 1.317, + "step": 339 + }, + { + "epoch": 0.3346456692913386, + "grad_norm": 17.13825035095215, + "learning_rate": 3.3399602385685887e-07, + "loss": 1.1972, + "step": 340 + }, + { + "epoch": 0.33562992125984253, + "grad_norm": 22.386058807373047, + "learning_rate": 3.349900596421472e-07, + "loss": 1.6323, + "step": 341 + }, + { + "epoch": 0.33661417322834647, + "grad_norm": 16.26036834716797, + "learning_rate": 3.359840954274355e-07, + "loss": 1.0469, + "step": 342 + }, + { + "epoch": 0.3375984251968504, + "grad_norm": 16.711883544921875, + "learning_rate": 3.369781312127237e-07, + "loss": 1.3349, + "step": 343 + }, + { + "epoch": 0.33858267716535434, + "grad_norm": 14.21274471282959, + "learning_rate": 3.37972166998012e-07, + "loss": 0.9544, + "step": 344 + }, + { + "epoch": 0.3395669291338583, + "grad_norm": 18.23252296447754, + "learning_rate": 3.3896620278330024e-07, + "loss": 1.1894, + "step": 345 + }, + { + "epoch": 0.3405511811023622, + "grad_norm": 13.04709243774414, + "learning_rate": 3.3996023856858855e-07, + "loss": 0.7717, + "step": 346 + }, + { + "epoch": 0.34153543307086615, + "grad_norm": 20.66889190673828, + "learning_rate": 3.409542743538768e-07, + "loss": 1.2563, + "step": 347 + }, + { + "epoch": 0.3425196850393701, + "grad_norm": 19.415496826171875, + "learning_rate": 3.4194831013916505e-07, + "loss": 1.2437, + "step": 348 + }, + { + "epoch": 0.343503937007874, + "grad_norm": 18.92940330505371, + "learning_rate": 3.4294234592445336e-07, + "loss": 0.7806, + "step": 349 + }, + { + "epoch": 0.34448818897637795, + "grad_norm": 18.86058235168457, + "learning_rate": 3.439363817097416e-07, + "loss": 0.8303, + "step": 350 + }, + { + "epoch": 0.3454724409448819, + "grad_norm": 22.05217933654785, + "learning_rate": 3.449304174950299e-07, + "loss": 1.0926, + "step": 351 + }, + { + "epoch": 0.3464566929133858, + "grad_norm": 18.41050910949707, + "learning_rate": 3.459244532803181e-07, + "loss": 0.6654, + "step": 352 + }, + { + "epoch": 0.34744094488188976, + "grad_norm": 20.6577091217041, + "learning_rate": 3.469184890656064e-07, + "loss": 1.1087, + "step": 353 + }, + { + "epoch": 0.3484251968503937, + "grad_norm": 21.558391571044922, + "learning_rate": 3.4791252485089467e-07, + "loss": 1.1525, + "step": 354 + }, + { + "epoch": 0.3494094488188976, + "grad_norm": 17.46851921081543, + "learning_rate": 3.48906560636183e-07, + "loss": 1.1127, + "step": 355 + }, + { + "epoch": 0.35039370078740156, + "grad_norm": 19.606983184814453, + "learning_rate": 3.499005964214713e-07, + "loss": 1.4267, + "step": 356 + }, + { + "epoch": 0.3513779527559055, + "grad_norm": 16.4560604095459, + "learning_rate": 3.508946322067595e-07, + "loss": 0.6148, + "step": 357 + }, + { + "epoch": 0.35236220472440943, + "grad_norm": 18.274593353271484, + "learning_rate": 3.518886679920478e-07, + "loss": 1.0123, + "step": 358 + }, + { + "epoch": 0.35334645669291337, + "grad_norm": 26.427845001220703, + "learning_rate": 3.5288270377733604e-07, + "loss": 1.9682, + "step": 359 + }, + { + "epoch": 0.3543307086614173, + "grad_norm": 15.270173072814941, + "learning_rate": 3.5387673956262434e-07, + "loss": 0.8487, + "step": 360 + }, + { + "epoch": 0.35531496062992124, + "grad_norm": 15.928024291992188, + "learning_rate": 3.5487077534791254e-07, + "loss": 1.0412, + "step": 361 + }, + { + "epoch": 0.3562992125984252, + "grad_norm": 16.021299362182617, + "learning_rate": 3.5586481113320085e-07, + "loss": 1.0902, + "step": 362 + }, + { + "epoch": 0.3572834645669291, + "grad_norm": 15.06041431427002, + "learning_rate": 3.5685884691848915e-07, + "loss": 0.9606, + "step": 363 + }, + { + "epoch": 0.35826771653543305, + "grad_norm": 17.597152709960938, + "learning_rate": 3.578528827037774e-07, + "loss": 0.9206, + "step": 364 + }, + { + "epoch": 0.359251968503937, + "grad_norm": 19.437129974365234, + "learning_rate": 3.5884691848906566e-07, + "loss": 1.4727, + "step": 365 + }, + { + "epoch": 0.36023622047244097, + "grad_norm": 16.03767204284668, + "learning_rate": 3.598409542743539e-07, + "loss": 0.9379, + "step": 366 + }, + { + "epoch": 0.3612204724409449, + "grad_norm": 16.83603286743164, + "learning_rate": 3.608349900596422e-07, + "loss": 0.8387, + "step": 367 + }, + { + "epoch": 0.36220472440944884, + "grad_norm": 18.7794132232666, + "learning_rate": 3.6182902584493047e-07, + "loss": 0.9692, + "step": 368 + }, + { + "epoch": 0.3631889763779528, + "grad_norm": 21.516565322875977, + "learning_rate": 3.6282306163021877e-07, + "loss": 1.6298, + "step": 369 + }, + { + "epoch": 0.3641732283464567, + "grad_norm": 14.46429443359375, + "learning_rate": 3.63817097415507e-07, + "loss": 1.0882, + "step": 370 + }, + { + "epoch": 0.36515748031496065, + "grad_norm": 18.343765258789062, + "learning_rate": 3.648111332007953e-07, + "loss": 1.1558, + "step": 371 + }, + { + "epoch": 0.3661417322834646, + "grad_norm": 16.055957794189453, + "learning_rate": 3.658051689860836e-07, + "loss": 0.9546, + "step": 372 + }, + { + "epoch": 0.3671259842519685, + "grad_norm": 15.39949893951416, + "learning_rate": 3.6679920477137184e-07, + "loss": 1.0124, + "step": 373 + }, + { + "epoch": 0.36811023622047245, + "grad_norm": 15.101058006286621, + "learning_rate": 3.677932405566601e-07, + "loss": 1.3916, + "step": 374 + }, + { + "epoch": 0.3690944881889764, + "grad_norm": 11.735250473022461, + "learning_rate": 3.6878727634194834e-07, + "loss": 0.527, + "step": 375 + }, + { + "epoch": 0.3700787401574803, + "grad_norm": 13.659126281738281, + "learning_rate": 3.6978131212723665e-07, + "loss": 0.6387, + "step": 376 + }, + { + "epoch": 0.37106299212598426, + "grad_norm": 18.6676025390625, + "learning_rate": 3.7077534791252495e-07, + "loss": 1.1445, + "step": 377 + }, + { + "epoch": 0.3720472440944882, + "grad_norm": 23.352149963378906, + "learning_rate": 3.7176938369781315e-07, + "loss": 1.3309, + "step": 378 + }, + { + "epoch": 0.37303149606299213, + "grad_norm": 25.64788818359375, + "learning_rate": 3.7276341948310145e-07, + "loss": 1.5888, + "step": 379 + }, + { + "epoch": 0.37401574803149606, + "grad_norm": 23.087392807006836, + "learning_rate": 3.737574552683897e-07, + "loss": 1.4422, + "step": 380 + }, + { + "epoch": 0.375, + "grad_norm": 21.51338005065918, + "learning_rate": 3.74751491053678e-07, + "loss": 1.7044, + "step": 381 + }, + { + "epoch": 0.37598425196850394, + "grad_norm": 13.316896438598633, + "learning_rate": 3.7574552683896626e-07, + "loss": 0.7913, + "step": 382 + }, + { + "epoch": 0.37696850393700787, + "grad_norm": 19.068180084228516, + "learning_rate": 3.767395626242545e-07, + "loss": 1.3241, + "step": 383 + }, + { + "epoch": 0.3779527559055118, + "grad_norm": 11.564128875732422, + "learning_rate": 3.777335984095428e-07, + "loss": 0.6473, + "step": 384 + }, + { + "epoch": 0.37893700787401574, + "grad_norm": 18.001808166503906, + "learning_rate": 3.787276341948311e-07, + "loss": 1.221, + "step": 385 + }, + { + "epoch": 0.3799212598425197, + "grad_norm": 13.824004173278809, + "learning_rate": 3.797216699801194e-07, + "loss": 0.7773, + "step": 386 + }, + { + "epoch": 0.3809055118110236, + "grad_norm": 14.364906311035156, + "learning_rate": 3.807157057654076e-07, + "loss": 1.054, + "step": 387 + }, + { + "epoch": 0.38188976377952755, + "grad_norm": 16.183347702026367, + "learning_rate": 3.817097415506959e-07, + "loss": 0.9862, + "step": 388 + }, + { + "epoch": 0.3828740157480315, + "grad_norm": 16.849651336669922, + "learning_rate": 3.8270377733598414e-07, + "loss": 0.9684, + "step": 389 + }, + { + "epoch": 0.3838582677165354, + "grad_norm": 17.902210235595703, + "learning_rate": 3.8369781312127244e-07, + "loss": 1.3244, + "step": 390 + }, + { + "epoch": 0.38484251968503935, + "grad_norm": 20.95703887939453, + "learning_rate": 3.8469184890656075e-07, + "loss": 1.1787, + "step": 391 + }, + { + "epoch": 0.3858267716535433, + "grad_norm": 22.145349502563477, + "learning_rate": 3.8568588469184895e-07, + "loss": 1.4698, + "step": 392 + }, + { + "epoch": 0.3868110236220472, + "grad_norm": 21.477500915527344, + "learning_rate": 3.8667992047713725e-07, + "loss": 1.0961, + "step": 393 + }, + { + "epoch": 0.38779527559055116, + "grad_norm": 17.14502716064453, + "learning_rate": 3.876739562624255e-07, + "loss": 1.1364, + "step": 394 + }, + { + "epoch": 0.3887795275590551, + "grad_norm": 18.223770141601562, + "learning_rate": 3.886679920477138e-07, + "loss": 0.9368, + "step": 395 + }, + { + "epoch": 0.38976377952755903, + "grad_norm": 20.13445472717285, + "learning_rate": 3.89662027833002e-07, + "loss": 1.1731, + "step": 396 + }, + { + "epoch": 0.390748031496063, + "grad_norm": 17.56426429748535, + "learning_rate": 3.906560636182903e-07, + "loss": 0.8686, + "step": 397 + }, + { + "epoch": 0.39173228346456695, + "grad_norm": 15.231781005859375, + "learning_rate": 3.916500994035786e-07, + "loss": 0.7481, + "step": 398 + }, + { + "epoch": 0.3927165354330709, + "grad_norm": 17.00189781188965, + "learning_rate": 3.9264413518886687e-07, + "loss": 0.7261, + "step": 399 + }, + { + "epoch": 0.3937007874015748, + "grad_norm": 17.566858291625977, + "learning_rate": 3.936381709741551e-07, + "loss": 1.2062, + "step": 400 + }, + { + "epoch": 0.39468503937007876, + "grad_norm": 15.011639595031738, + "learning_rate": 3.946322067594434e-07, + "loss": 0.7462, + "step": 401 + }, + { + "epoch": 0.3956692913385827, + "grad_norm": 19.905258178710938, + "learning_rate": 3.956262425447317e-07, + "loss": 1.0318, + "step": 402 + }, + { + "epoch": 0.39665354330708663, + "grad_norm": 17.95429229736328, + "learning_rate": 3.9662027833001993e-07, + "loss": 1.105, + "step": 403 + }, + { + "epoch": 0.39763779527559057, + "grad_norm": 18.217300415039062, + "learning_rate": 3.9761431411530824e-07, + "loss": 1.009, + "step": 404 + }, + { + "epoch": 0.3986220472440945, + "grad_norm": 13.397478103637695, + "learning_rate": 3.986083499005965e-07, + "loss": 0.5941, + "step": 405 + }, + { + "epoch": 0.39960629921259844, + "grad_norm": 21.784292221069336, + "learning_rate": 3.9960238568588474e-07, + "loss": 1.7972, + "step": 406 + }, + { + "epoch": 0.4005905511811024, + "grad_norm": 21.588985443115234, + "learning_rate": 4.0059642147117305e-07, + "loss": 1.0544, + "step": 407 + }, + { + "epoch": 0.4015748031496063, + "grad_norm": 18.62330436706543, + "learning_rate": 4.015904572564613e-07, + "loss": 1.3912, + "step": 408 + }, + { + "epoch": 0.40255905511811024, + "grad_norm": 14.669224739074707, + "learning_rate": 4.0258449304174955e-07, + "loss": 0.8305, + "step": 409 + }, + { + "epoch": 0.4035433070866142, + "grad_norm": 15.314811706542969, + "learning_rate": 4.035785288270378e-07, + "loss": 0.8688, + "step": 410 + }, + { + "epoch": 0.4045275590551181, + "grad_norm": 17.41103172302246, + "learning_rate": 4.045725646123261e-07, + "loss": 1.0069, + "step": 411 + }, + { + "epoch": 0.40551181102362205, + "grad_norm": 18.243221282958984, + "learning_rate": 4.055666003976144e-07, + "loss": 1.3141, + "step": 412 + }, + { + "epoch": 0.406496062992126, + "grad_norm": 14.436189651489258, + "learning_rate": 4.065606361829026e-07, + "loss": 1.1042, + "step": 413 + }, + { + "epoch": 0.4074803149606299, + "grad_norm": 18.251113891601562, + "learning_rate": 4.075546719681909e-07, + "loss": 1.1011, + "step": 414 + }, + { + "epoch": 0.40846456692913385, + "grad_norm": 18.68360710144043, + "learning_rate": 4.085487077534792e-07, + "loss": 1.1192, + "step": 415 + }, + { + "epoch": 0.4094488188976378, + "grad_norm": 23.807735443115234, + "learning_rate": 4.095427435387675e-07, + "loss": 1.5957, + "step": 416 + }, + { + "epoch": 0.4104330708661417, + "grad_norm": 17.419734954833984, + "learning_rate": 4.1053677932405573e-07, + "loss": 1.164, + "step": 417 + }, + { + "epoch": 0.41141732283464566, + "grad_norm": 13.43248462677002, + "learning_rate": 4.11530815109344e-07, + "loss": 0.6425, + "step": 418 + }, + { + "epoch": 0.4124015748031496, + "grad_norm": 11.816337585449219, + "learning_rate": 4.125248508946323e-07, + "loss": 0.6068, + "step": 419 + }, + { + "epoch": 0.41338582677165353, + "grad_norm": 16.29872703552246, + "learning_rate": 4.1351888667992054e-07, + "loss": 0.9275, + "step": 420 + }, + { + "epoch": 0.41437007874015747, + "grad_norm": 16.47966957092285, + "learning_rate": 4.1451292246520885e-07, + "loss": 0.8836, + "step": 421 + }, + { + "epoch": 0.4153543307086614, + "grad_norm": 18.686519622802734, + "learning_rate": 4.1550695825049705e-07, + "loss": 1.2115, + "step": 422 + }, + { + "epoch": 0.41633858267716534, + "grad_norm": 17.829578399658203, + "learning_rate": 4.1650099403578535e-07, + "loss": 0.8367, + "step": 423 + }, + { + "epoch": 0.41732283464566927, + "grad_norm": 17.083707809448242, + "learning_rate": 4.174950298210736e-07, + "loss": 1.0595, + "step": 424 + }, + { + "epoch": 0.4183070866141732, + "grad_norm": 17.380428314208984, + "learning_rate": 4.184890656063619e-07, + "loss": 0.826, + "step": 425 + }, + { + "epoch": 0.41929133858267714, + "grad_norm": 14.618813514709473, + "learning_rate": 4.194831013916502e-07, + "loss": 0.707, + "step": 426 + }, + { + "epoch": 0.4202755905511811, + "grad_norm": 13.27505111694336, + "learning_rate": 4.204771371769384e-07, + "loss": 0.6235, + "step": 427 + }, + { + "epoch": 0.421259842519685, + "grad_norm": 17.46605682373047, + "learning_rate": 4.214711729622267e-07, + "loss": 0.7719, + "step": 428 + }, + { + "epoch": 0.422244094488189, + "grad_norm": 19.197139739990234, + "learning_rate": 4.2246520874751497e-07, + "loss": 1.0862, + "step": 429 + }, + { + "epoch": 0.42322834645669294, + "grad_norm": 19.14566421508789, + "learning_rate": 4.234592445328033e-07, + "loss": 0.9311, + "step": 430 + }, + { + "epoch": 0.4242125984251969, + "grad_norm": 23.42700958251953, + "learning_rate": 4.244532803180915e-07, + "loss": 1.2339, + "step": 431 + }, + { + "epoch": 0.4251968503937008, + "grad_norm": 19.94841766357422, + "learning_rate": 4.254473161033798e-07, + "loss": 0.9891, + "step": 432 + }, + { + "epoch": 0.42618110236220474, + "grad_norm": 22.30583953857422, + "learning_rate": 4.264413518886681e-07, + "loss": 1.8443, + "step": 433 + }, + { + "epoch": 0.4271653543307087, + "grad_norm": 19.260499954223633, + "learning_rate": 4.2743538767395634e-07, + "loss": 1.1799, + "step": 434 + }, + { + "epoch": 0.4281496062992126, + "grad_norm": 16.3272762298584, + "learning_rate": 4.284294234592446e-07, + "loss": 0.759, + "step": 435 + }, + { + "epoch": 0.42913385826771655, + "grad_norm": 17.50089454650879, + "learning_rate": 4.2942345924453284e-07, + "loss": 1.1002, + "step": 436 + }, + { + "epoch": 0.4301181102362205, + "grad_norm": 14.367104530334473, + "learning_rate": 4.3041749502982115e-07, + "loss": 0.9141, + "step": 437 + }, + { + "epoch": 0.4311023622047244, + "grad_norm": 11.757801055908203, + "learning_rate": 4.3141153081510945e-07, + "loss": 0.5467, + "step": 438 + }, + { + "epoch": 0.43208661417322836, + "grad_norm": 16.255727767944336, + "learning_rate": 4.324055666003977e-07, + "loss": 0.7476, + "step": 439 + }, + { + "epoch": 0.4330708661417323, + "grad_norm": 20.517709732055664, + "learning_rate": 4.3339960238568596e-07, + "loss": 1.14, + "step": 440 + }, + { + "epoch": 0.4340551181102362, + "grad_norm": 18.24642562866211, + "learning_rate": 4.343936381709742e-07, + "loss": 1.1504, + "step": 441 + }, + { + "epoch": 0.43503937007874016, + "grad_norm": 23.892919540405273, + "learning_rate": 4.353876739562625e-07, + "loss": 1.26, + "step": 442 + }, + { + "epoch": 0.4360236220472441, + "grad_norm": 17.60642433166504, + "learning_rate": 4.3638170974155077e-07, + "loss": 1.0311, + "step": 443 + }, + { + "epoch": 0.43700787401574803, + "grad_norm": 18.21918487548828, + "learning_rate": 4.37375745526839e-07, + "loss": 1.0646, + "step": 444 + }, + { + "epoch": 0.43799212598425197, + "grad_norm": 17.054752349853516, + "learning_rate": 4.383697813121273e-07, + "loss": 0.8687, + "step": 445 + }, + { + "epoch": 0.4389763779527559, + "grad_norm": 15.23765754699707, + "learning_rate": 4.393638170974156e-07, + "loss": 0.6839, + "step": 446 + }, + { + "epoch": 0.43996062992125984, + "grad_norm": 17.00933837890625, + "learning_rate": 4.403578528827039e-07, + "loss": 1.1376, + "step": 447 + }, + { + "epoch": 0.4409448818897638, + "grad_norm": 15.645007133483887, + "learning_rate": 4.413518886679921e-07, + "loss": 0.9759, + "step": 448 + }, + { + "epoch": 0.4419291338582677, + "grad_norm": 17.53651237487793, + "learning_rate": 4.423459244532804e-07, + "loss": 0.7971, + "step": 449 + }, + { + "epoch": 0.44291338582677164, + "grad_norm": 19.482398986816406, + "learning_rate": 4.4333996023856864e-07, + "loss": 0.9708, + "step": 450 + }, + { + "epoch": 0.4438976377952756, + "grad_norm": 14.216598510742188, + "learning_rate": 4.4433399602385695e-07, + "loss": 0.8217, + "step": 451 + }, + { + "epoch": 0.4448818897637795, + "grad_norm": 18.865297317504883, + "learning_rate": 4.4532803180914525e-07, + "loss": 1.3728, + "step": 452 + }, + { + "epoch": 0.44586614173228345, + "grad_norm": 16.348478317260742, + "learning_rate": 4.4632206759443345e-07, + "loss": 0.9119, + "step": 453 + }, + { + "epoch": 0.4468503937007874, + "grad_norm": 18.980161666870117, + "learning_rate": 4.4731610337972176e-07, + "loss": 1.012, + "step": 454 + }, + { + "epoch": 0.4478346456692913, + "grad_norm": 20.558147430419922, + "learning_rate": 4.4831013916501e-07, + "loss": 1.3738, + "step": 455 + }, + { + "epoch": 0.44881889763779526, + "grad_norm": 16.78011703491211, + "learning_rate": 4.493041749502983e-07, + "loss": 0.8219, + "step": 456 + }, + { + "epoch": 0.4498031496062992, + "grad_norm": 19.52912712097168, + "learning_rate": 4.502982107355865e-07, + "loss": 1.2558, + "step": 457 + }, + { + "epoch": 0.4507874015748031, + "grad_norm": 15.149005889892578, + "learning_rate": 4.512922465208748e-07, + "loss": 0.6247, + "step": 458 + }, + { + "epoch": 0.45177165354330706, + "grad_norm": 18.364900588989258, + "learning_rate": 4.522862823061631e-07, + "loss": 0.7295, + "step": 459 + }, + { + "epoch": 0.45177165354330706, + "eval_Qnli-dev_cosine_accuracy": 0.693359375, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8319265842437744, + "eval_Qnli-dev_cosine_ap": 0.6873625888187367, + "eval_Qnli-dev_cosine_f1": 0.685337726523888, + "eval_Qnli-dev_cosine_f1_threshold": 0.74552983045578, + "eval_Qnli-dev_cosine_precision": 0.5606469002695418, + "eval_Qnli-dev_cosine_recall": 0.8813559322033898, + "eval_Qnli-dev_dot_accuracy": 0.693359375, + "eval_Qnli-dev_dot_accuracy_threshold": 639.0776977539062, + "eval_Qnli-dev_dot_ap": 0.6878718449643791, + "eval_Qnli-dev_dot_f1": 0.685337726523888, + "eval_Qnli-dev_dot_f1_threshold": 572.7136840820312, + "eval_Qnli-dev_dot_precision": 0.5606469002695418, + "eval_Qnli-dev_dot_recall": 0.8813559322033898, + "eval_Qnli-dev_euclidean_accuracy": 0.693359375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 16.06937026977539, + "eval_Qnli-dev_euclidean_ap": 0.6873686687008952, + "eval_Qnli-dev_euclidean_f1": 0.685337726523888, + "eval_Qnli-dev_euclidean_f1_threshold": 19.772865295410156, + "eval_Qnli-dev_euclidean_precision": 0.5606469002695418, + "eval_Qnli-dev_euclidean_recall": 0.8813559322033898, + "eval_Qnli-dev_manhattan_accuracy": 0.69921875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 362.1485900878906, + "eval_Qnli-dev_manhattan_ap": 0.6874910715870401, + "eval_Qnli-dev_manhattan_f1": 0.6857142857142857, + "eval_Qnli-dev_manhattan_f1_threshold": 430.38519287109375, + "eval_Qnli-dev_manhattan_precision": 0.5682451253481894, + "eval_Qnli-dev_manhattan_recall": 0.864406779661017, + "eval_Qnli-dev_max_accuracy": 0.69921875, + "eval_Qnli-dev_max_accuracy_threshold": 639.0776977539062, + "eval_Qnli-dev_max_ap": 0.6878718449643791, + "eval_Qnli-dev_max_f1": 0.6857142857142857, + "eval_Qnli-dev_max_f1_threshold": 572.7136840820312, + "eval_Qnli-dev_max_precision": 0.5682451253481894, + "eval_Qnli-dev_max_recall": 0.8813559322033898, + "eval_allNLI-dev_cosine_accuracy": 0.68359375, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9088386297225952, + "eval_allNLI-dev_cosine_ap": 0.48873606015680937, + "eval_allNLI-dev_cosine_f1": 0.5350553505535056, + "eval_allNLI-dev_cosine_f1_threshold": 0.8140230178833008, + "eval_allNLI-dev_cosine_precision": 0.39295392953929537, + "eval_allNLI-dev_cosine_recall": 0.838150289017341, + "eval_allNLI-dev_dot_accuracy": 0.68359375, + "eval_allNLI-dev_dot_accuracy_threshold": 699.0950927734375, + "eval_allNLI-dev_dot_ap": 0.48885724782911755, + "eval_allNLI-dev_dot_f1": 0.5350553505535056, + "eval_allNLI-dev_dot_f1_threshold": 625.3240356445312, + "eval_allNLI-dev_dot_precision": 0.39295392953929537, + "eval_allNLI-dev_dot_recall": 0.838150289017341, + "eval_allNLI-dev_euclidean_accuracy": 0.68359375, + "eval_allNLI-dev_euclidean_accuracy_threshold": 11.834823608398438, + "eval_allNLI-dev_euclidean_ap": 0.4887203371983184, + "eval_allNLI-dev_euclidean_f1": 0.5350553505535056, + "eval_allNLI-dev_euclidean_f1_threshold": 16.90357780456543, + "eval_allNLI-dev_euclidean_precision": 0.39295392953929537, + "eval_allNLI-dev_euclidean_recall": 0.838150289017341, + "eval_allNLI-dev_manhattan_accuracy": 0.68359375, + "eval_allNLI-dev_manhattan_accuracy_threshold": 256.45477294921875, + "eval_allNLI-dev_manhattan_ap": 0.4920209563997524, + "eval_allNLI-dev_manhattan_f1": 0.5396145610278372, + "eval_allNLI-dev_manhattan_f1_threshold": 339.225830078125, + "eval_allNLI-dev_manhattan_precision": 0.42857142857142855, + "eval_allNLI-dev_manhattan_recall": 0.7283236994219653, + "eval_allNLI-dev_max_accuracy": 0.68359375, + "eval_allNLI-dev_max_accuracy_threshold": 699.0950927734375, + "eval_allNLI-dev_max_ap": 0.4920209563997524, + "eval_allNLI-dev_max_f1": 0.5396145610278372, + "eval_allNLI-dev_max_f1_threshold": 625.3240356445312, + "eval_allNLI-dev_max_precision": 0.42857142857142855, + "eval_allNLI-dev_max_recall": 0.838150289017341, + "eval_runtime": 1.2229, + "eval_samples_per_second": 0.0, + "eval_sequential_score": 0.6878718449643791, + "eval_steps_per_second": 0.0, + "eval_sts-test_pearson_cosine": 0.5377382226514003, + "eval_sts-test_pearson_dot": 0.5376496659087263, + "eval_sts-test_pearson_euclidean": 0.5469897917607747, + "eval_sts-test_pearson_manhattan": 0.5464293120330461, + "eval_sts-test_pearson_max": 0.5469897917607747, + "eval_sts-test_spearman_cosine": 0.5410237309359288, + "eval_sts-test_spearman_dot": 0.5408811086658744, + "eval_sts-test_spearman_euclidean": 0.5409800984560722, + "eval_sts-test_spearman_manhattan": 0.5401021234588343, + "eval_sts-test_spearman_max": 0.5410237309359288, + "step": 459 + }, + { + "epoch": 0.452755905511811, + "grad_norm": 17.561704635620117, + "learning_rate": 4.532803180914514e-07, + "loss": 0.8154, + "step": 460 + }, + { + "epoch": 0.453740157480315, + "grad_norm": 21.252965927124023, + "learning_rate": 4.542743538767397e-07, + "loss": 1.1392, + "step": 461 + }, + { + "epoch": 0.4547244094488189, + "grad_norm": 18.453880310058594, + "learning_rate": 4.552683896620279e-07, + "loss": 0.8618, + "step": 462 + }, + { + "epoch": 0.45570866141732286, + "grad_norm": 17.269075393676758, + "learning_rate": 4.562624254473162e-07, + "loss": 0.9669, + "step": 463 + }, + { + "epoch": 0.4566929133858268, + "grad_norm": 20.30898094177246, + "learning_rate": 4.5725646123260444e-07, + "loss": 0.8804, + "step": 464 + }, + { + "epoch": 0.45767716535433073, + "grad_norm": 17.70629119873047, + "learning_rate": 4.5825049701789274e-07, + "loss": 0.8479, + "step": 465 + }, + { + "epoch": 0.45866141732283466, + "grad_norm": 17.85552215576172, + "learning_rate": 4.59244532803181e-07, + "loss": 0.6296, + "step": 466 + }, + { + "epoch": 0.4596456692913386, + "grad_norm": 20.20493507385254, + "learning_rate": 4.6023856858846925e-07, + "loss": 0.8449, + "step": 467 + }, + { + "epoch": 0.46062992125984253, + "grad_norm": 17.319948196411133, + "learning_rate": 4.6123260437375755e-07, + "loss": 0.9772, + "step": 468 + }, + { + "epoch": 0.46161417322834647, + "grad_norm": 14.453144073486328, + "learning_rate": 4.622266401590458e-07, + "loss": 0.6424, + "step": 469 + }, + { + "epoch": 0.4625984251968504, + "grad_norm": 19.411174774169922, + "learning_rate": 4.6322067594433406e-07, + "loss": 0.9169, + "step": 470 + }, + { + "epoch": 0.46358267716535434, + "grad_norm": 15.519804954528809, + "learning_rate": 4.642147117296223e-07, + "loss": 0.7599, + "step": 471 + }, + { + "epoch": 0.4645669291338583, + "grad_norm": 18.444625854492188, + "learning_rate": 4.652087475149106e-07, + "loss": 0.8943, + "step": 472 + }, + { + "epoch": 0.4655511811023622, + "grad_norm": 15.110048294067383, + "learning_rate": 4.662027833001989e-07, + "loss": 0.9475, + "step": 473 + }, + { + "epoch": 0.46653543307086615, + "grad_norm": 21.10722541809082, + "learning_rate": 4.671968190854872e-07, + "loss": 1.4518, + "step": 474 + }, + { + "epoch": 0.4675196850393701, + "grad_norm": 20.026628494262695, + "learning_rate": 4.681908548707754e-07, + "loss": 1.274, + "step": 475 + }, + { + "epoch": 0.468503937007874, + "grad_norm": 14.985383987426758, + "learning_rate": 4.691848906560637e-07, + "loss": 0.7306, + "step": 476 + }, + { + "epoch": 0.46948818897637795, + "grad_norm": 17.787303924560547, + "learning_rate": 4.70178926441352e-07, + "loss": 0.9238, + "step": 477 + }, + { + "epoch": 0.4704724409448819, + "grad_norm": 12.555164337158203, + "learning_rate": 4.7117296222664024e-07, + "loss": 0.6593, + "step": 478 + }, + { + "epoch": 0.4714566929133858, + "grad_norm": 17.015419006347656, + "learning_rate": 4.721669980119285e-07, + "loss": 1.0183, + "step": 479 + }, + { + "epoch": 0.47244094488188976, + "grad_norm": 18.000446319580078, + "learning_rate": 4.731610337972168e-07, + "loss": 1.2577, + "step": 480 + }, + { + "epoch": 0.4734251968503937, + "grad_norm": 17.278921127319336, + "learning_rate": 4.7415506958250505e-07, + "loss": 0.8738, + "step": 481 + }, + { + "epoch": 0.4744094488188976, + "grad_norm": 17.930349349975586, + "learning_rate": 4.7514910536779335e-07, + "loss": 1.1416, + "step": 482 + }, + { + "epoch": 0.47539370078740156, + "grad_norm": 13.973302841186523, + "learning_rate": 4.7614314115308155e-07, + "loss": 0.7135, + "step": 483 + }, + { + "epoch": 0.4763779527559055, + "grad_norm": 18.695253372192383, + "learning_rate": 4.771371769383698e-07, + "loss": 1.2587, + "step": 484 + }, + { + "epoch": 0.47736220472440943, + "grad_norm": 13.285400390625, + "learning_rate": 4.781312127236581e-07, + "loss": 0.8823, + "step": 485 + }, + { + "epoch": 0.47834645669291337, + "grad_norm": 15.625781059265137, + "learning_rate": 4.791252485089464e-07, + "loss": 0.8423, + "step": 486 + }, + { + "epoch": 0.4793307086614173, + "grad_norm": 15.811039924621582, + "learning_rate": 4.801192842942347e-07, + "loss": 0.7704, + "step": 487 + }, + { + "epoch": 0.48031496062992124, + "grad_norm": 17.256977081298828, + "learning_rate": 4.811133200795229e-07, + "loss": 0.7049, + "step": 488 + }, + { + "epoch": 0.4812992125984252, + "grad_norm": 20.418750762939453, + "learning_rate": 4.821073558648112e-07, + "loss": 1.1893, + "step": 489 + }, + { + "epoch": 0.4822834645669291, + "grad_norm": 21.882749557495117, + "learning_rate": 4.831013916500995e-07, + "loss": 1.3985, + "step": 490 + }, + { + "epoch": 0.48326771653543305, + "grad_norm": 21.558752059936523, + "learning_rate": 4.840954274353877e-07, + "loss": 1.3567, + "step": 491 + }, + { + "epoch": 0.484251968503937, + "grad_norm": 23.495107650756836, + "learning_rate": 4.85089463220676e-07, + "loss": 1.2573, + "step": 492 + }, + { + "epoch": 0.48523622047244097, + "grad_norm": 14.981607437133789, + "learning_rate": 4.860834990059642e-07, + "loss": 0.7671, + "step": 493 + }, + { + "epoch": 0.4862204724409449, + "grad_norm": 17.21627426147461, + "learning_rate": 4.870775347912525e-07, + "loss": 0.5425, + "step": 494 + }, + { + "epoch": 0.48720472440944884, + "grad_norm": 20.916128158569336, + "learning_rate": 4.880715705765408e-07, + "loss": 0.9372, + "step": 495 + }, + { + "epoch": 0.4881889763779528, + "grad_norm": 15.274470329284668, + "learning_rate": 4.890656063618291e-07, + "loss": 0.799, + "step": 496 + }, + { + "epoch": 0.4891732283464567, + "grad_norm": 18.584993362426758, + "learning_rate": 4.900596421471173e-07, + "loss": 0.9548, + "step": 497 + }, + { + "epoch": 0.49015748031496065, + "grad_norm": 18.775136947631836, + "learning_rate": 4.910536779324057e-07, + "loss": 1.0855, + "step": 498 + }, + { + "epoch": 0.4911417322834646, + "grad_norm": 24.180419921875, + "learning_rate": 4.92047713717694e-07, + "loss": 1.0465, + "step": 499 + }, + { + "epoch": 0.4921259842519685, + "grad_norm": 18.497556686401367, + "learning_rate": 4.930417495029823e-07, + "loss": 1.1004, + "step": 500 + }, + { + "epoch": 0.49311023622047245, + "grad_norm": 14.301048278808594, + "learning_rate": 4.940357852882705e-07, + "loss": 0.6392, + "step": 501 + }, + { + "epoch": 0.4940944881889764, + "grad_norm": 14.273008346557617, + "learning_rate": 4.950298210735588e-07, + "loss": 0.7102, + "step": 502 + }, + { + "epoch": 0.4950787401574803, + "grad_norm": 18.05272102355957, + "learning_rate": 4.96023856858847e-07, + "loss": 1.3242, + "step": 503 + }, + { + "epoch": 0.49606299212598426, + "grad_norm": 13.768959045410156, + "learning_rate": 4.970178926441353e-07, + "loss": 0.6861, + "step": 504 + }, + { + "epoch": 0.4970472440944882, + "grad_norm": 14.028704643249512, + "learning_rate": 4.980119284294235e-07, + "loss": 0.9291, + "step": 505 + }, + { + "epoch": 0.49803149606299213, + "grad_norm": 14.870085716247559, + "learning_rate": 4.990059642147118e-07, + "loss": 0.8592, + "step": 506 + }, + { + "epoch": 0.49901574803149606, + "grad_norm": 15.56464672088623, + "learning_rate": 5.000000000000001e-07, + "loss": 0.9462, + "step": 507 + }, + { + "epoch": 0.5, + "grad_norm": 15.741423606872559, + "learning_rate": 5.009940357852883e-07, + "loss": 1.0167, + "step": 508 + }, + { + "epoch": 0.5009842519685039, + "grad_norm": 18.02831268310547, + "learning_rate": 5.019880715705767e-07, + "loss": 1.0118, + "step": 509 + }, + { + "epoch": 0.5019685039370079, + "grad_norm": 15.293427467346191, + "learning_rate": 5.029821073558649e-07, + "loss": 0.6741, + "step": 510 + }, + { + "epoch": 0.5029527559055118, + "grad_norm": 19.62667465209961, + "learning_rate": 5.039761431411532e-07, + "loss": 1.4578, + "step": 511 + }, + { + "epoch": 0.5039370078740157, + "grad_norm": 20.828519821166992, + "learning_rate": 5.049701789264414e-07, + "loss": 1.2959, + "step": 512 + }, + { + "epoch": 0.5049212598425197, + "grad_norm": 14.50117015838623, + "learning_rate": 5.059642147117297e-07, + "loss": 0.8533, + "step": 513 + }, + { + "epoch": 0.5059055118110236, + "grad_norm": 15.186738967895508, + "learning_rate": 5.06958250497018e-07, + "loss": 0.6685, + "step": 514 + }, + { + "epoch": 0.5068897637795275, + "grad_norm": 17.811168670654297, + "learning_rate": 5.079522862823062e-07, + "loss": 1.1556, + "step": 515 + }, + { + "epoch": 0.5078740157480315, + "grad_norm": 17.15772247314453, + "learning_rate": 5.089463220675945e-07, + "loss": 0.8177, + "step": 516 + }, + { + "epoch": 0.5088582677165354, + "grad_norm": 12.814756393432617, + "learning_rate": 5.099403578528828e-07, + "loss": 0.6296, + "step": 517 + }, + { + "epoch": 0.5098425196850394, + "grad_norm": 14.35216999053955, + "learning_rate": 5.10934393638171e-07, + "loss": 0.8407, + "step": 518 + }, + { + "epoch": 0.5108267716535433, + "grad_norm": 14.875484466552734, + "learning_rate": 5.119284294234593e-07, + "loss": 0.6987, + "step": 519 + }, + { + "epoch": 0.5118110236220472, + "grad_norm": 15.191465377807617, + "learning_rate": 5.129224652087476e-07, + "loss": 0.9888, + "step": 520 + }, + { + "epoch": 0.5127952755905512, + "grad_norm": 15.37837028503418, + "learning_rate": 5.139165009940358e-07, + "loss": 0.8938, + "step": 521 + }, + { + "epoch": 0.5137795275590551, + "grad_norm": 13.611766815185547, + "learning_rate": 5.149105367793241e-07, + "loss": 0.582, + "step": 522 + }, + { + "epoch": 0.514763779527559, + "grad_norm": 17.264280319213867, + "learning_rate": 5.159045725646124e-07, + "loss": 0.6596, + "step": 523 + }, + { + "epoch": 0.515748031496063, + "grad_norm": 15.050413131713867, + "learning_rate": 5.168986083499007e-07, + "loss": 0.6029, + "step": 524 + }, + { + "epoch": 0.5167322834645669, + "grad_norm": 19.290075302124023, + "learning_rate": 5.178926441351889e-07, + "loss": 0.9806, + "step": 525 + }, + { + "epoch": 0.5177165354330708, + "grad_norm": 15.048296928405762, + "learning_rate": 5.188866799204771e-07, + "loss": 0.9463, + "step": 526 + }, + { + "epoch": 0.5187007874015748, + "grad_norm": 15.788368225097656, + "learning_rate": 5.198807157057656e-07, + "loss": 0.7088, + "step": 527 + }, + { + "epoch": 0.5196850393700787, + "grad_norm": 15.568428039550781, + "learning_rate": 5.208747514910538e-07, + "loss": 0.7525, + "step": 528 + }, + { + "epoch": 0.5206692913385826, + "grad_norm": 13.9373197555542, + "learning_rate": 5.21868787276342e-07, + "loss": 0.7625, + "step": 529 + }, + { + "epoch": 0.5216535433070866, + "grad_norm": 16.379072189331055, + "learning_rate": 5.228628230616304e-07, + "loss": 0.8271, + "step": 530 + }, + { + "epoch": 0.5226377952755905, + "grad_norm": 17.776195526123047, + "learning_rate": 5.238568588469186e-07, + "loss": 0.6129, + "step": 531 + }, + { + "epoch": 0.5236220472440944, + "grad_norm": 24.5928897857666, + "learning_rate": 5.248508946322069e-07, + "loss": 1.1563, + "step": 532 + }, + { + "epoch": 0.5246062992125984, + "grad_norm": 17.05551528930664, + "learning_rate": 5.258449304174951e-07, + "loss": 0.8131, + "step": 533 + }, + { + "epoch": 0.5255905511811023, + "grad_norm": 14.337945938110352, + "learning_rate": 5.268389662027834e-07, + "loss": 0.5363, + "step": 534 + }, + { + "epoch": 0.5265748031496063, + "grad_norm": 20.1900634765625, + "learning_rate": 5.278330019880717e-07, + "loss": 0.8819, + "step": 535 + }, + { + "epoch": 0.5275590551181102, + "grad_norm": 24.67251205444336, + "learning_rate": 5.288270377733599e-07, + "loss": 0.9772, + "step": 536 + }, + { + "epoch": 0.5285433070866141, + "grad_norm": 22.329355239868164, + "learning_rate": 5.298210735586483e-07, + "loss": 1.2102, + "step": 537 + }, + { + "epoch": 0.5295275590551181, + "grad_norm": 22.19415855407715, + "learning_rate": 5.308151093439365e-07, + "loss": 1.1234, + "step": 538 + }, + { + "epoch": 0.530511811023622, + "grad_norm": 23.826335906982422, + "learning_rate": 5.318091451292247e-07, + "loss": 1.1857, + "step": 539 + }, + { + "epoch": 0.531496062992126, + "grad_norm": 16.586576461791992, + "learning_rate": 5.32803180914513e-07, + "loss": 0.7873, + "step": 540 + }, + { + "epoch": 0.53248031496063, + "grad_norm": 12.48427677154541, + "learning_rate": 5.337972166998013e-07, + "loss": 0.5034, + "step": 541 + }, + { + "epoch": 0.5334645669291339, + "grad_norm": 20.41586685180664, + "learning_rate": 5.347912524850895e-07, + "loss": 1.3305, + "step": 542 + }, + { + "epoch": 0.5344488188976378, + "grad_norm": 22.552961349487305, + "learning_rate": 5.357852882703778e-07, + "loss": 1.1727, + "step": 543 + }, + { + "epoch": 0.5354330708661418, + "grad_norm": 21.429880142211914, + "learning_rate": 5.367793240556661e-07, + "loss": 1.2825, + "step": 544 + }, + { + "epoch": 0.5364173228346457, + "grad_norm": 19.160961151123047, + "learning_rate": 5.377733598409544e-07, + "loss": 1.0446, + "step": 545 + }, + { + "epoch": 0.5374015748031497, + "grad_norm": 17.107141494750977, + "learning_rate": 5.387673956262426e-07, + "loss": 0.9838, + "step": 546 + }, + { + "epoch": 0.5383858267716536, + "grad_norm": 18.039453506469727, + "learning_rate": 5.397614314115308e-07, + "loss": 1.2194, + "step": 547 + }, + { + "epoch": 0.5393700787401575, + "grad_norm": 13.434783935546875, + "learning_rate": 5.407554671968192e-07, + "loss": 0.7709, + "step": 548 + }, + { + "epoch": 0.5403543307086615, + "grad_norm": 14.461050987243652, + "learning_rate": 5.417495029821074e-07, + "loss": 0.748, + "step": 549 + }, + { + "epoch": 0.5413385826771654, + "grad_norm": 18.0071964263916, + "learning_rate": 5.427435387673957e-07, + "loss": 1.0948, + "step": 550 + }, + { + "epoch": 0.5423228346456693, + "grad_norm": 15.916511535644531, + "learning_rate": 5.43737574552684e-07, + "loss": 0.915, + "step": 551 + }, + { + "epoch": 0.5433070866141733, + "grad_norm": 22.280460357666016, + "learning_rate": 5.447316103379722e-07, + "loss": 1.537, + "step": 552 + }, + { + "epoch": 0.5442913385826772, + "grad_norm": 9.682397842407227, + "learning_rate": 5.457256461232605e-07, + "loss": 0.3239, + "step": 553 + }, + { + "epoch": 0.5452755905511811, + "grad_norm": 14.292939186096191, + "learning_rate": 5.467196819085487e-07, + "loss": 0.9592, + "step": 554 + }, + { + "epoch": 0.5462598425196851, + "grad_norm": 22.54702377319336, + "learning_rate": 5.477137176938371e-07, + "loss": 0.7737, + "step": 555 + }, + { + "epoch": 0.547244094488189, + "grad_norm": 14.62065315246582, + "learning_rate": 5.487077534791253e-07, + "loss": 0.613, + "step": 556 + }, + { + "epoch": 0.5482283464566929, + "grad_norm": 22.537317276000977, + "learning_rate": 5.497017892644135e-07, + "loss": 1.3646, + "step": 557 + }, + { + "epoch": 0.5492125984251969, + "grad_norm": 15.06298542022705, + "learning_rate": 5.50695825049702e-07, + "loss": 0.6659, + "step": 558 + }, + { + "epoch": 0.5501968503937008, + "grad_norm": 13.530902862548828, + "learning_rate": 5.516898608349902e-07, + "loss": 0.5207, + "step": 559 + }, + { + "epoch": 0.5511811023622047, + "grad_norm": 19.478429794311523, + "learning_rate": 5.526838966202784e-07, + "loss": 0.9467, + "step": 560 + }, + { + "epoch": 0.5521653543307087, + "grad_norm": 10.947720527648926, + "learning_rate": 5.536779324055667e-07, + "loss": 0.5692, + "step": 561 + }, + { + "epoch": 0.5531496062992126, + "grad_norm": 33.76875686645508, + "learning_rate": 5.54671968190855e-07, + "loss": 1.5855, + "step": 562 + }, + { + "epoch": 0.5541338582677166, + "grad_norm": 18.264493942260742, + "learning_rate": 5.556660039761433e-07, + "loss": 0.8855, + "step": 563 + }, + { + "epoch": 0.5551181102362205, + "grad_norm": 25.19763946533203, + "learning_rate": 5.566600397614315e-07, + "loss": 1.1829, + "step": 564 + }, + { + "epoch": 0.5561023622047244, + "grad_norm": 22.214569091796875, + "learning_rate": 5.576540755467198e-07, + "loss": 0.978, + "step": 565 + }, + { + "epoch": 0.5570866141732284, + "grad_norm": 15.043540000915527, + "learning_rate": 5.586481113320081e-07, + "loss": 1.1818, + "step": 566 + }, + { + "epoch": 0.5580708661417323, + "grad_norm": 15.76466178894043, + "learning_rate": 5.596421471172963e-07, + "loss": 0.701, + "step": 567 + }, + { + "epoch": 0.5590551181102362, + "grad_norm": 16.75844955444336, + "learning_rate": 5.606361829025846e-07, + "loss": 1.0226, + "step": 568 + }, + { + "epoch": 0.5600393700787402, + "grad_norm": 10.848587989807129, + "learning_rate": 5.616302186878729e-07, + "loss": 0.5937, + "step": 569 + }, + { + "epoch": 0.5610236220472441, + "grad_norm": 15.039306640625, + "learning_rate": 5.626242544731611e-07, + "loss": 0.8095, + "step": 570 + }, + { + "epoch": 0.562007874015748, + "grad_norm": 17.843490600585938, + "learning_rate": 5.636182902584494e-07, + "loss": 1.174, + "step": 571 + }, + { + "epoch": 0.562992125984252, + "grad_norm": 15.040814399719238, + "learning_rate": 5.646123260437377e-07, + "loss": 0.96, + "step": 572 + }, + { + "epoch": 0.5639763779527559, + "grad_norm": 17.238296508789062, + "learning_rate": 5.656063618290259e-07, + "loss": 0.8339, + "step": 573 + }, + { + "epoch": 0.5649606299212598, + "grad_norm": 13.707849502563477, + "learning_rate": 5.666003976143142e-07, + "loss": 0.717, + "step": 574 + }, + { + "epoch": 0.5659448818897638, + "grad_norm": 14.237223625183105, + "learning_rate": 5.675944333996024e-07, + "loss": 0.5938, + "step": 575 + }, + { + "epoch": 0.5669291338582677, + "grad_norm": 12.927653312683105, + "learning_rate": 5.685884691848908e-07, + "loss": 0.6501, + "step": 576 + }, + { + "epoch": 0.5679133858267716, + "grad_norm": 15.383443832397461, + "learning_rate": 5.69582504970179e-07, + "loss": 0.7003, + "step": 577 + }, + { + "epoch": 0.5688976377952756, + "grad_norm": 14.056544303894043, + "learning_rate": 5.705765407554672e-07, + "loss": 0.5525, + "step": 578 + }, + { + "epoch": 0.5698818897637795, + "grad_norm": 17.352069854736328, + "learning_rate": 5.715705765407556e-07, + "loss": 0.7003, + "step": 579 + }, + { + "epoch": 0.5708661417322834, + "grad_norm": 22.5910587310791, + "learning_rate": 5.725646123260438e-07, + "loss": 1.059, + "step": 580 + }, + { + "epoch": 0.5718503937007874, + "grad_norm": 19.49325942993164, + "learning_rate": 5.735586481113321e-07, + "loss": 0.8625, + "step": 581 + }, + { + "epoch": 0.5728346456692913, + "grad_norm": 14.92470645904541, + "learning_rate": 5.745526838966203e-07, + "loss": 0.5862, + "step": 582 + }, + { + "epoch": 0.5738188976377953, + "grad_norm": 16.41839027404785, + "learning_rate": 5.755467196819086e-07, + "loss": 0.9162, + "step": 583 + }, + { + "epoch": 0.5748031496062992, + "grad_norm": 20.332698822021484, + "learning_rate": 5.765407554671969e-07, + "loss": 0.926, + "step": 584 + }, + { + "epoch": 0.5757874015748031, + "grad_norm": 22.728179931640625, + "learning_rate": 5.775347912524851e-07, + "loss": 1.2729, + "step": 585 + }, + { + "epoch": 0.5767716535433071, + "grad_norm": 16.058364868164062, + "learning_rate": 5.785288270377734e-07, + "loss": 0.8935, + "step": 586 + }, + { + "epoch": 0.577755905511811, + "grad_norm": 14.835709571838379, + "learning_rate": 5.795228628230618e-07, + "loss": 0.541, + "step": 587 + }, + { + "epoch": 0.5787401574803149, + "grad_norm": 19.8588809967041, + "learning_rate": 5.8051689860835e-07, + "loss": 1.1455, + "step": 588 + }, + { + "epoch": 0.5797244094488189, + "grad_norm": 14.425910949707031, + "learning_rate": 5.815109343936383e-07, + "loss": 0.7306, + "step": 589 + }, + { + "epoch": 0.5807086614173228, + "grad_norm": 17.056396484375, + "learning_rate": 5.825049701789266e-07, + "loss": 0.9088, + "step": 590 + }, + { + "epoch": 0.5816929133858267, + "grad_norm": 14.791122436523438, + "learning_rate": 5.834990059642148e-07, + "loss": 0.9166, + "step": 591 + }, + { + "epoch": 0.5826771653543307, + "grad_norm": 14.80226993560791, + "learning_rate": 5.844930417495031e-07, + "loss": 0.8679, + "step": 592 + }, + { + "epoch": 0.5836614173228346, + "grad_norm": 17.1209716796875, + "learning_rate": 5.854870775347914e-07, + "loss": 0.9329, + "step": 593 + }, + { + "epoch": 0.5846456692913385, + "grad_norm": 17.293729782104492, + "learning_rate": 5.864811133200797e-07, + "loss": 1.1201, + "step": 594 + }, + { + "epoch": 0.5856299212598425, + "grad_norm": 15.181586265563965, + "learning_rate": 5.874751491053679e-07, + "loss": 0.6418, + "step": 595 + }, + { + "epoch": 0.5866141732283464, + "grad_norm": 18.143863677978516, + "learning_rate": 5.884691848906561e-07, + "loss": 1.145, + "step": 596 + }, + { + "epoch": 0.5875984251968503, + "grad_norm": 19.976341247558594, + "learning_rate": 5.894632206759445e-07, + "loss": 1.4041, + "step": 597 + }, + { + "epoch": 0.5885826771653543, + "grad_norm": 13.67269515991211, + "learning_rate": 5.904572564612327e-07, + "loss": 0.6954, + "step": 598 + }, + { + "epoch": 0.5895669291338582, + "grad_norm": 14.636340141296387, + "learning_rate": 5.914512922465209e-07, + "loss": 0.4567, + "step": 599 + }, + { + "epoch": 0.5905511811023622, + "grad_norm": 16.68194580078125, + "learning_rate": 5.924453280318093e-07, + "loss": 1.1305, + "step": 600 + }, + { + "epoch": 0.5915354330708661, + "grad_norm": 17.648611068725586, + "learning_rate": 5.934393638170975e-07, + "loss": 0.8077, + "step": 601 + }, + { + "epoch": 0.59251968503937, + "grad_norm": 14.298154830932617, + "learning_rate": 5.944333996023858e-07, + "loss": 0.6143, + "step": 602 + }, + { + "epoch": 0.593503937007874, + "grad_norm": 19.324697494506836, + "learning_rate": 5.95427435387674e-07, + "loss": 1.3139, + "step": 603 + }, + { + "epoch": 0.594488188976378, + "grad_norm": 14.576811790466309, + "learning_rate": 5.964214711729623e-07, + "loss": 0.7694, + "step": 604 + }, + { + "epoch": 0.5954724409448819, + "grad_norm": 15.335044860839844, + "learning_rate": 5.974155069582506e-07, + "loss": 0.9622, + "step": 605 + }, + { + "epoch": 0.5964566929133859, + "grad_norm": 16.537038803100586, + "learning_rate": 5.984095427435388e-07, + "loss": 0.91, + "step": 606 + }, + { + "epoch": 0.5974409448818898, + "grad_norm": 21.826786041259766, + "learning_rate": 5.994035785288272e-07, + "loss": 1.3125, + "step": 607 + }, + { + "epoch": 0.5984251968503937, + "grad_norm": 13.646114349365234, + "learning_rate": 6.003976143141154e-07, + "loss": 1.0153, + "step": 608 + }, + { + "epoch": 0.5994094488188977, + "grad_norm": 11.858122825622559, + "learning_rate": 6.013916500994036e-07, + "loss": 0.8468, + "step": 609 + }, + { + "epoch": 0.6003937007874016, + "grad_norm": 16.532007217407227, + "learning_rate": 6.02385685884692e-07, + "loss": 1.1026, + "step": 610 + } + ], + "logging_steps": 1, + "max_steps": 3048, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 305, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 32, + "trial_name": null, + "trial_params": null +}